Harvey Sacks Lectures On Conversation, Volumes I, II The Estate Of Harvey Sacks ( 1995)

LECTURES ON
CONVERSATION

VOLUMES I & II

Harvey Sacks

Edited

by

Gail Jefferson

With an Introduction by

Emanuel A. Schegloff

Blackwell

Publishing

Lectures on Conversation, Volume I, II Harvey Sacks
© 1995 The Estate of Harvey Sacks. ISBN: 978-1-557-86705-6

© 1992, 1995 by The Estate of Harvey Sacks
Introduction © 1992, 1995 by Emanuel A. Schegloff

BLACKWELL PUBLISHING

350 Main Street, Malden, MA 02148-5020, USA

9600 Garsington Road, Oxford OX4 2DQ, UK

550 Swanston Street, Carlton, Victoria 3053, Australia

All rights reserved. No part of this publication may be reproduced, stored in a
retrieval system, or transmitted, in any form or by any means, electronic,
mechanical, photocopying, recording or otherwise, except as permitted by the UK
Copyright, Designs, and Patents Act 1988, without the prior permission of the
publisher.

First published as two volumes 1992

First published as one paperback volume 1995

5 2006

Library of Congress Cataloging-in-Publication Data

Harvey Sacks : lectures on conversation / edited by Gail Jefferson; with an
introduction by Emanuel A. Schegloff.
p. cm.

Includes bibliographical references and index.

ISBN 1-55786-705-4 (pb)

1 . Conversation. 2. Oral communication. I. Jefferson, Gail.

P95 .45 .H37 1992

302.3' 46— dc20 91-45536

CIP

ISBN-13: 978-1-55786-705-6 (pb)

A catalogue record for this title is available from the British Library.

Printed and bound in the United Kingdom
by TJ International Ltd, Padstow, Cornwall

The publisher's policy is to use permanent paper from mills that operate a
sustainable forestry policy, and which has been manufactured from pulp
processed using acid-free and elementary chlorine-free practices. Furthermore,
the publisher ensures that the text paper and cover board used have met
acceptable environmental accreditation standards.

For further information on

Blackwell Publishing, visit our website:

www.blackwellpublishing.com

LECTURES ON
CONVERSATION

VOLUME I

LECTURES ON
CONVERSATION

VOLUME II

Harvey Sacks, c. 1967

Harvey Sacks, c. 1968

Contents

Introduction by Emanuel A. Schegloff ix

Note lxiii

Acknowledgments lxv

Part I Fall 1964-Spring 1965

Lecture 1 Rules of conversational sequence 3

Lecture 2 On suicide threats getting laughed off 12

Lecture 3 The correction-invitation device 2 1

Lecture 4 An impromptu survey of the literature 26

Lecture 3 Suicide as a device for discovering if anybody cares 32

Lecture 6 The MIR membership categorization device 40

Lecture 7 On questions 49

Lecture 8 On measuring 57

Lecture 9 “I am nothing” 66

Lecture 10 Accountable actions 72

Lecture 1 1 On exchanging glances 8 1

Lecture 12 Sequencing: Utterances, jokes, and questions 95

Lecture 13 On proverbs 104

Lecture 1 4 The inference-making machine 113

Appendix A A Note on the Editing 126

Part II Fall 1965

[Lecture 1 and 2] [‘‘The baby cried. The mommy picked it up.”] 135

Handout Group therapy session segment 136

Lecture 3 A collaboratively built sentence; The use of ‘We’ 144

Lecture 4 Tying rules 150

Lecture 3 Tying rules; Insult sequences 157

Lecture 6 ‘You’ 163

Lecture 1 ‘Hotrodders’ as a revolutionary category 169

Lecture 8 Invitations; Inexhaustable topics; Category-bound

activities 175

Lecture 9 Character appears on cue; Good grounds for

an action 182

Lecture 10 Clausal construction; Hotrodding as a test 188

Lecture 1 1 Espousing a rule; Exemplary occurrences 193

VI

Contents

Lecture 12 ‘Tearing down;’ Non-translatable categories 199

{Lecture 13] [‘Everyone has to lie’] 204

Lecture 14 The navy pilot [from Sacks’ Research Notes] 205

Appendix A “The baby cried” [Notes for lecture 1] 223

Appendix B “The baby cried” [Notes on lecture 2] 230

Part III Spring 1966

Lecture 1 “The baby cried. The mommy picked it up” 236

Lecture 1(R) “The baby cried. The mommy picked it up” 243

Lecture 2 “The baby cried. The mommy picked it up” (ctd) 252

Lecture 2(R) “The baby cried. The mommy picked it up” (ctd) 259
[Lecture 3} [‘Everyone has to lie’] 267

Handout Group therapy session segment 268

Lecture 04.a An introduction sequence 281

Lecture 04.b An introduction sequence (ctd) 292

Lecture 4 Invitations; Identifications; Category-bound activities 300

Lecture 5 Proffering identifications; The navy pilot; Slots;

Paired objects, Adequate complete utterances 306

Lecture 6 Omni-relevant devices; Cover identifications 312

Lecture 7 Cover topics; Collaborative sentences; Tying

rules; Relational-pair identifications 320

Lecture 08 Orientation; Being ‘phoney;’ Hinting 328

Lecture 8 ‘We;’ Category-bound activities 333

{Lecture 9] [incorporated into lectures 8 and 10] 341

Lecture 10 Pro-verbs; Performatives; Position markers;

Warnings 342

Lecture 11 ‘You’ 348

Lecture 12 Warnings, challenges, and corrections;

Explanations; Complaining-praising; Games 354

Lecture 13 Button-button who’s got the button 363

Lecture 14 Disorderability; Tying rules 370

Lecture 1 5 Tying rules; Playing dumb; Correction-invitation

device 376

Lecture 16 Possessive pronouns; Possessables and possessitives 382

Lecture 17 Pervasive, inexhaustable topics; Emblems 389

Lecture 18 ‘Hotrodders’ as a revolutionary category 396

Lecture 19 Appearance verbs 404

Lecture 20 Character appears on cue; Good grounds for an action;

An explanation is the explanation 410

Lecture 21 Misidentification; Membership categories; Utterance

pairs; Paradoxes 417

/ Lecture 22] [combined with lecture 21] 427

Lecture 23 Agreement; What can be done with language? 428

Lecture 24 Measurement systems 435

Lecture 23 ‘Company’ as an alternation category [incomplete] 44 1

Contents

Lecture 26 Being ‘chicken’ versus ‘giving lip back’ 443

Lecture 21 A mis-hearing (“a green?”); A taboo on hearing 450

Lecture 28 Intelligibility; Causally efficacious categories 456

Lecture 29 Place references; Weak and safe compliments 46 1

Lecture 30 Various methodological issues 467

Lecture 31 Games: legal and illegal actions 473

Lecture 32 Seeing an ‘imitation’ 479

Lecture 33 On sampling and subjectivity 483

Appendix A ‘On some formal properties of children’s games’ 489

Appendix B A Note on the Editing 507

Part IV Winter 1967

February 16 Omnirelevant devices; Settinged activities; ‘Indicator

terms’ 515

March 2 Turn-taking; Collaborative utterances via appendor

questions; Instructions; Directed utterances 523

March 9 Topic; Utterance placement; ‘Activity occupied’

phenomena; Formulations; Euphemisms 535

Part V Spring 1967

[Lecture 1—7} {‘One party speaks at a time’]

Lecture 8 ‘‘Everyone has to lie” 549

Lecture 9 ‘‘Everyone has to lie”(ctd) 557

[Lecture 10] {incorporated into lectures 9 and 11] 567

Lecture 1 1 ‘We;’ Category-bound activities; ‘Stereotypes’ 568

Lecture 12 Category-bound activities; Programmatic relevance;

Hinting; Being ‘phoney’ 578

Lecture 13 Category-bound activities: ‘‘The baby cried;” Praising,

warning, and challenging; Tautological proverbs 584

Lecture 14 ‘Cover’ categories; Omni-relevance 590

Lecture 13.1 ‘Safe’ compliments and complaints 597

Lecture 13.2 Ultra-rich topics 601

Lecture 16 Possessables and possessitives 605

Lecture 1 7 Claiming possession; Emblems; Pro-terms and

performatives; Utterance positioning 610

[Lecture 18] {Paradoxes] 616

Part VI Fall 1967

General Introduction 619

Lecture 1 The speaker sequencing problem 624

Lecture 2 The ‘one-at-a-time’ rule; Violations; Complaints;

Gossip 633

Lecture 3 Bases for ‘interruption’ 64 1

Contents

viii

Lecture 4 Utterance completion; Co-producing an utterance;

Appendor clauses 647

Lecture 5 Utterance completion; Action sequencing; Appendor

questions 656

Lecture 6 Next-speaker selection techniques; Paired utterances 665

Lecture 7 Intentional mis-address; Floor seekers 675

Lecture 8 Pre-sequences 685

Lecture 9 Paradoxes 693

Lecture 10 ‘Everyone gets a chance to talk’ 701

Lecture Oil Pronouns 711

Lecture 11 Tying techniques 716

Lecture 12 Repetition tying; “A green?” . . . “Who’s Wayne

Morse?” 722

Lecture 13 Tying-based mis-hearings; Locational tying; Pro-verbs

and performatives 730

Lecture 14 Paraphrasing; Alternative temporal references;

Approximate and precise numbers; Laughter; ‘Uh huh’ 739

Part VII Spring 1968

April 17 Topic 752

April 24 Second Stories 764

May 8 Reason for a call; Tellability 773

May 22 Pauses in spelling and numbering 784

May 29 Verb selection; Interactionally generated invitations;

Adequacy of local environments, etc. 787

Appendix I ‘Introduction’ 1965 802

Bibliography 806

Index 813

Contents

Introduction by Emanuel A. Schegloff ix

Part I Fall 1968

Lecture 1

Second stories; “Mm hm;” Story prefaces; ‘Local news;’
Tellability

3

Lecture 2

Features of a recognizable ‘story;’ Story prefaces;
Sequential locator terms; Lawful interruption

17

Lecture 3

Turn-taking; The notion ‘conversation;’ Noticeable
absences; Greetings; Adjacency

32

Lecture 4

Turn-taking; Complaints about interruption;
Enforcement

44

Lecture 3

Collaboratives; Possible utterances; Utterance pairs;
Greetings and introductions

56

Lecture 6

Greetings and introductions; Orientational utterances;
Ultra rich, infinite topics; Being ‘phoney’

67

Part II

Winter 1969

Lecture 1

Announcements; Touched-off utterances; Noticings;
The makings of conversation; Local resources

87

Lecture 2

Safe Compliments

98

Lecture 3

‘Patients with observers’ as 'performers with audience’

104

Lecture 7

Alternative sequences; Challenges; Claiming
membership

114

Lecture 8

‘Identification reformulation;’ Pairing off at parties;
‘Abstract’ versus ‘concrete’ formulations

126

Lecture 9

Sound shifts; Showing understanding; Dealing with
‘utterance completion;’ Practical mysticism

137

Fragment

Verb uses; ‘A puzzle about pronouns’

150

Part III

Winter 1970

Lecture 1

Foreshortened versus expanded greeting sequences;
Voice recognition tests; Reason for a call; ‘My mind
is with you;’ Tellability

157

Lecture 2

Conveying information; Story-connective techniques;
Recognition-type descriptors; ‘First verbs;’
Understanding; Differential organization of perception

175

VI

Contents

Lecture 4 Greetings: Adjacency pairs; Sequential implicativeness;

The integrative function of public tragedy 188

Lecture 5 Foreshortened, normal, and expanded beginning
sequences; Joking relationships; First topics; Close
offerings 200

Part IV Spring 1970

Lecture 1 Doing ‘being ordinary’ 215

Lecture 2 Stories take more than one utterance; Story prefaces 222

Lecture 3 Story organization; Tellability; Coincidence, etc. 229

Lecture 4 Storyteller as ‘witness;’ Entitlement to experience 242

Lecture 5 ‘First’ and ‘second’ stories; Topical coherence; Storing

and recalling experiences 249

Lecture 6 Hypothetical second stories and explanations for first
stories; Sound-related terms (Poetics); “What I didn’t
do” 261

Lecture 7 ‘What’s going on’ in a lay sense; Tracking

co-participants; Context information; Pre-positioned
laughter; Interpreting utterances not directed to one 269

Lecture 8 Asking questions; Heckling 282

Part V Winter 1971

February 19 Poetics; Tracking co-participants; Touched-off topics,

Stepwise topical movement 291

March 4 Produced similarities in first and second stories;

Poetics; ‘Fragile stories,’ etc. 303

March 1 1 Poetics; Requests, offers, and threats; The ‘old man’

as an evolved natural object 318

Part VI Spring 1971

April 2 Introduction 335

April 5 Poetics; Avoiding speaking first 340

April 9 Technical competition 348

April 12 Long sequences 354

April 19 Caller-Called 360

April 23 Characterizing an event 367

April 26 An event as an institution 370

April 30 Calling for help 376

May 3 Problem solving; Recipient-designed solutions 384

May 10 Agent-client interaction 391

May 1 7 Poetics: Spatialized characterizations 396

May 21 Closing; Communicating a feeling; Doctor as

‘stranger’ 402

May 24 “Uh huh;” Questioner-preferred answers 410

Contents vii

Part VII Fall 1971

Lecture 1 On hypothetical data; Puns; Proverbial expressions 419

Lecture 2 Doing ‘understanding;’ Puns 425

Lecture 3 Allusive talk; Poetics 431

Lecture 4 Spouse talk 437

Lecture 3 Selecting identifications 444

Lecture 6 A ‘defensively designed’ story 453

Lecture 7 The ‘motive power’ of a story; ‘Ex-relationals’ 458

Lecture 8 Preserving and transmitting knowledge via stories 466

Lecture 9 The dirty joke as a technical object; Temporal and

sequential organization; ‘Guiding’ recipient 470

Lecture 10 The dirty joke as a technical object (ctd); Suspending

disbelief; ‘Guiding’ recipient; Punchlines 478

Lecture 1 1 The dirty joke as a technical object (ctd); Packaging

and transmitting experiences 483

Lecture 12 The dirty joke as a technical object (ctd); “What is sex

like;” Possible versus actual applicability of a rule 489

Lecture 13 Two ‘floor-seizure’ techniques: Appositional expletives

and “Uh” 495

Lecture 14 The working of a list; Doing ‘hostility’ 499

Lecture 13 ‘Fragile’ stories; On being ‘rational’ 504

Lecture 16 On dreams 512

Part VIII Spring 1972

Lecture 1 Adjacency pairs: Scope of operation 521

Lecture 2 Adjacency pairs: Distribution in conversation; A single

instance of a Q-A pair 533

Lecture 3 A single instance of a phone-call opening;

Caller-Called, etc. 542

Lecture 4 The relating power of adjacency; Next position 554

Lecture 3 A single instance of a Q-A pair; Topical versus pair

organization; Disaster talk 561

Lecture (6) Laughing together; Expressions of sorrow and joy 570

References 576

Index

577

Introduction

The publication of these lectures makes publicly available virtually all of the
lectures by Harvey Sacks on conversation and related topics in social science.
Most of the lectures in this larger corpus were originally delivered to classes at
the University of California - first to sociology classes at the UCLA campus,
and then (beginning in Fall 1968) to classes in the School of Social Science at
the Irvine campus of the University.

Although Sacks produced copious analytic notes, many of which served as
materials for his lectures, what is presented here are the lectures themselves,
transcribed from tape recordings. Almost all of Sacks’ lectures were initially
transcribed by Gail Jefferson, although most of the material for Fall
1964-Spring 1965, in that it antedates either her contact with Sacks and this
work, or her undertaking to transcribe the lectures, was intially transcribed by
others. With one exception (Sacks, 1987 {1973}), it is also Jefferson who has
edited those lectures which have previously been published, as well as the
lectures published here. 1 As noted in her introductory notes to the several

My thanks to Paul Drew, John Heritage, Gail Jefferson, Michael Moerman and
Melvin Pollner for sensitive responses to a draft of an earlier version of part of this
introduction (prepared for the 1989 publication of the 1964-5 lectures), and for
suggestions which I have in some cases adopted without further acknowledgement.
I am further indebted to John Heritage and Michael Moerman for their generous and
helpful comments on a draft of the present introduction/memoir, and to Gail
Jefferson for calling to my attention what she took to be lapses in accuracy or
taste.

1 Of the lectures published here, the set for 1964-5 were published in a special
issue of the journal Human Studies , 12, 3-4 (1989), and of those, the following had
been previously published elsewhere, edited by Gail Jefferson:

Fall 1964-5, lecture 5 has been published under the title ‘You want to find out
if anybody really does care’ in Button and Lee (1987: 217-25).

Winter 1964-5, lecture 14 has been published under the title ‘The inference
making machine: notes on observability’ in van Dijk (1985: 13-22).

Other than the 1964-5 lectures, the following lectures have been previously
published, also edited by Jefferson:

Spring 1966, lecture 18 (and related material in Fall 1965, lecture 7), under the
title ‘Hotrodder: a revolutionary category,’ in Psathas (1979).

Spring 1966, lecture 13, under the title ‘Button-button who’s got the button,’ in
Zimmerman and West (1980: 318-27).

Spring 1966, lecture 24 (with excerpts from Fall 1967, lecture 14; Winter 1970,
lecture 2; and Spring 1970, lecture 3), under the title ‘On members’ measurement
systems,’ (Sacks, 1988/89).

tx

X

Introduction

‘lectures’ and in Appendix II in her editor’s notes to the previous publication
of the 1964-5 lectures in the journal Human Studies , those lecture-texts have
been pieced together from several sets of lectures which Sacks gave during the
1964-5 academic year, to make a more coherent and readable document.
These early ‘sets’ of lectures are full of gaps, and it is not always clear just
when some lecture was given. Accordingly, the reader should bear in mind
that this presentation of Sacks’ early lectures cannot be used to track the
development of themes over time, to trace what topics or themes appear to
have been related in Sacks’ thinking, etc.

Otherwise, it needs to be said at the outset with respect to the present
edition that the editorial undertaking has been monumental and its execution
heroic. This the reader can only partially see, for what has not been included
is, for that reason, not apparent, nor is the work of sorting and collating what
is made available in these volumes. This work has, as a matter of course,
involved divergences of several kinds from the texts of these lectures which
have circulated in various forms of reproduction over the years. These are
largely stylistic in nature, and are clearly designed to render the text more
accessible, more readable, and more consistent in stance, point of view, diction,
etc. 2 On occasion, however, these textual adjustments could be misread as
taking a stand on an analytic matter which Sacks otherwise addresses, could be
given a ‘political’ reading, or could appear to have a ‘political’ upshot, and it
would be well for the reader to be alerted to such possibilities.

By ‘taking a stand on an analytic matter which Sacks otherwise addresses’
I mean to call attention to such adjustments in diction as one in which Sacks
follows an excerpt from a group therapy session by referring to one of the
speakers as “this fellow Dan” (in the originally circulated transcript of the
lecture), a reference which is in the present edition rendered as ‘the therapist.’
Sacks takes up the issue of the description of persons, and category-ascriptions
such as ‘therapist/patient,’ on several different occasions in these lectures and
in several papers. Because of the options available for formulating persons,
particular choices of descriptors or identification terms served, in Sacks’ view,
to pose problems for analysis, and could not properly be invoked or employed
in an unexamined way. Accordingly, no particular claims should be under-

Winter 1970, lectures 1 and 2, under the title ‘Some considerations of a story told
in ordinary conversations,’ (Sacks, 1986).

Spring 1970, lecture 1 (with excerpts from Winter 1970, lecture 2; Spring 1970,
lecture 4; and Spring 1971, lecture 1), under the title ‘On doing being ordinary,’ in
Atkinson and Heritage (1984).

Fall 1971, lectures 9-12, under the title ‘Some technical considerations of a dirty
joke,’ in Schenkein (1978).

In addition, extracts from a number of lectures have been assembled by Jefferson
as ‘Notes on methodology,’ in Atkinson and Heritage (1984: 21-7).

2 Cf. the editor’s notes by Gail Jefferson (Jefferson, 1989), and, in the present
edition, footnotes at Fall 1964-Spring 1965, lecture 2, p. 18; Spring 1966, lecture
04. a, p. 281; Winter 1969, lecture 7, p. 120; as well as Appendix A to lectures for
Fall 1964-Spring 1965, and Appendix B to lectures for Spring 1966.

Introduction

xi

stood as implied by occasional references to participants by such category
terms in the current text (cf. the editor’s Appendix A, Spring 1966).

By ‘political’ I mean, in this context, a relative positioning by Sacks of
himself, his undertaking, his colleagues, his students, other contemporary
intellectual undertakings, the established contours of the disciplines (sociol-
ogy, linguistics, anthropology, etc.) and their groupings (e.g., the social
sciences), the physically present class to which he was ostensibly addressing
himself 3 and the like. Deployment of the pronouns ‘we,’ ‘you,’ ‘they’ and the
like can serve to express varying sorts of solidarity and differentiation, and
different ways of ‘partitioning the population’ (as he used to put it). 4 This was
a matter to which Sacks was sensitive, having written a paper in graduate
school only a few years earlier on Durkheim’s use of ‘we’ in The Elementary
Forms of the Religious Life, an echo of which appears in lecture 33 of the
Spring 1966 lectures. Where the text suggests such alignments, readers
should exercise caution.

It must also be recalled that the omission of some lecture sets in the present
edition, and the transposition of some lectures from one set to another,
requires that caution be used in basing an analysis of the appearance and
development of themes, ideas and discussions of data fragments on this
edition alone. The full texts of prior versions of all the lectures will be
available through the Sacks Archive at the Department of Special Collections
of the UCLA Library.

These cautions aside, it should be said that one cannot really retrieve Sacks’
‘voice’ from the text as presented here. In the interest of readability and of the
accessibility of the content, what was sometimes a real challenge to discursive
parsing - even to his closest friends and colleagues - has been smoothed out.
Gone are the often convoluted phrasing, the syntax that might or might not
come together at the end, the often apparently pointillistic movement from
observation to observation - sometimes dovetailing at the end into a coherent
argument or picture, sometimes not. The very long silences, of course, were
lost in the transcribing process.

But Sacks himself treated his habits and manners, his attitudes and
convictions, as ‘private’ (as he puts it in response to a question as to whether
he is ‘convinced’ that single events are studyable after the general introductory
lecture, Fall 1967, “That’s such a private question”), and not really relevant

3 Cf. the lecture of April 2 in the Spring 1971 lectures, on Sacks’ notion that he
was really talking to colleagues, friends and ‘students’ wherever they might be who
were interested in his current work and not necessarily to the class actually in the
room.

1 For example, in lecture 6 of the 1964-5 lectures an alignment may seem to be
implied in which Sacks identifies himself with the physically present students in
criticism of “all the sociology u>e read,” whereas the text of the lecture as previously
circulated had read “all the sociology you read ...” (emphases supplied).

See Appendix A to the Fall 1964-Spring 1965 lectures for the editor’s account of
Sacks’ use of personal pronouns such as ‘you,’ ‘I’ and ‘we’ in the lectures, and of her
editorial practices for changing some of these references in preparing this edition.

Introduction

xii

to what he felt merited the attention of others in what he had to say. It is that
which these volumes present. As quickly becomes apparent from the texts of
the lectures, we have yet to take the full measure of the man.

These series of lectures present a most remarkable, inventive and produc-
tive account of a strikingly new vision of how to study human sociality. With
but a few exceptions, the students who sat in the rooms in which the lectures
were delivered can hardly have known what they were hearing. The lectures
were addressed to non-present students, to those who might come to know
what to make of them. That audience continues to grow.

Under what circumstances were these lectures delivered and recorded?
What is their intellectual and scientific context? What is most notable in
them? These matters cannot be dealt with comprehensively here, but a brief
treatment, in a mixed genre which might be termed an ‘introduction/
memoir,’ can help provide an overview and some setting for what is
increasingly recognized as a startlingly original and important address to the
social organization of mind, culture and interaction.

/

Sacks received his AB degree in 1955 after three years at Columbia College.
In later years, Sacks would reminisce with partly feigned and partly genuine
awe about the faculty at Columbia - Jacques Barzun, Meyer Schapiro, Lionel
Trilling, various students and former students of Franz Neumann such as
Julian Franklin and Peter Gay (and Neumann himself, who, however, may
well have not been teaching undergraduates when Sacks was there), although
it was never entirely clear with which of these ‘eminences’ Sacks had himself
studied. 5

Although he did not officially ‘major’ in sociology, Sacks’ education was
influenced in an important way by C. Wright Mills. The influence was not
channeled primarily through course work; most important to Sacks was that
Mills secured for him a faculty-authorized access to the stacks of Butler
Library and turned him loose on his own. But Sacks would later say that from
Mills he had learned ‘audacity.’

In spite of the predominantly socio-cultural cast of the faculty who figured
most centrally in Sacks’ later reminiscences, the two closest college friends
with whom Sacks kept in touch later on were both biologists.

Upon graduation from Columbia, Sacks was awarded a scholarship at Yale
Law School where he earned his LLB in 1959- While at Yale, he participated
in the group around Harold Lasswell, and became more interested in
understanding how the law as an institution worked, how it could work, than

5 I recall an account of how students would celebrate if they achieved grades of ‘A’
from Trilling or Schapiro, but it was unclear, at least to me, whether Sacks himself
had been one of those students.

Introduction

xiii

in making it work as an attorney himself. 6 He went looking for intellectual
resources with which to pursue this interest and turned first to Cambridge,
and to the work of Talcott Parsons in particular (although formally he was
enrolled as a graduate student in Political Science at MIT, and was employed
as a research assistant in the Department of Economics and Social Sciences
there). But what he found in Cambridge that was most consequential for the
subsequent development of his thinking was not Parsons (and not Chomsky,
some of whose lectures at MIT he attended), but rather Harold Garfinkel.

Garfinkel was spending a sabbatical leave from UCLA at Harvard, where
he had himself earned his Ph D a number of years earlier. Sacks and Garfinkel
met at Parsons’ seminar in Cambridge, and were immediately attracted by
each other’s seriousness. Their intellectual relationship was sustained until the
early 1970s. However, in 1959-60, when it became clear to Sacks that the
solutions to the problems he had set himself were not to be found in
Cambridge, he followed his law school teacher Lasswell’s advice, and decided
to pursue graduate work in sociology at Berkeley.

Berkeley appealed on several grounds. Laswell had suggested that Sacks
pursue his interests through the continuing study of labor law and industrial
relations. An attractive locale was furnished by the Institute of Industrial

6 Sacks once recounted a story which provides some insight into the appeal which
Garfinkel’s work must have had for him when he later encountered it.

He was engaged in a discussion with several other law school students arguing
through some problem in case law which they had been set - a problem in torts, if
I remember correctly. The issue was whether or not a person on the ground was
entitled to recover damages incurred from the overflight of his property by an
airplane. At one point, in a kind of mimicry of the ‘how many hairs make a bald
man’ paradox, the students coped with the argument that no damages could be
collected if the plane was being piloted in a proper and accepted manner by seeing
how far they could press the definition of what was ‘proper.’ What if it were flying
at 2,000 feet? At 1,000 feet? At 250 feet? At 5 feet? Sacks reported that when the
last of these proposals was offered, it was dismissed as ‘unreasonable,’ as frivolous, as
violating the canons of ‘common sense.’ But, he pointed out, that could as well have
been said about the penultimate one, but wasn’t. What struck him, then, and
puzzled him, was that the 'legal reasoning’ which was the much heralded instrument
in whose use they (students of the law) were being trained rested on, and was
constrained by, an infrastructure of so-called ‘common sense’ which was entirely tacit
and beyond the reach of argument, while controlling it. And, in that legal reasoning
was something on which the entire legal structure rested (and not just particular areas,
such as torts, contracts, crimes, etc.), how the law as an institution actually worked,
what made it work the way it did, what restrained its reasoning from pressing the law
in other directions, was shrouded in mystery. Undoubtedly, this was only one of the
puzzles about how the law could work which engaged Sacks’ interest, but it is one
for the solution of which Garfinkel’s work on methods of commonsense reasoning
and practical theorizing, then in progress, would have been an attractive resource.

The issue prompted by this law school incident gets articulated explicitly for its
bearing on working with recorded conversational materials at the beginning of lecture
1 for Fall 1971; cf. volume 2 of the present edition.

XIV

Introduction

Relations at Berkeley, and in particular by Philip Selznick whose interest in
organizations and bureaucracy was complemented by a developing interest in
legal institutions. (Indeed, several years later Selznick was to establish the
Center for the Study of Law and Society at Berkeley, and Sacks was to be
among its first graduate fellows.) But Berkeley was attractive on other counts
as well. Aside from its having developed one of the strongest sociology
departments in the country, Sacks was attracted by the presence of Herbert
Blumer, whose SSRC monographic critique (1939) of Thomas and
Znaniecki’s Polish Peasant in Europe and America Sacks had found penetrat-
ing. (Sacks lost interest in Blumer soon after arriving in Berkeley, and did not
study with him at all.)

It is worth pausing a moment to recall where some of the relevant
American social sciences stood during these formative years of the late 1950s
and early 1960s, at least as they appeared to graduate students, to some
graduate students, to the graduate students who figure in this account.

There had not yet been the rise to professional visibility of a radical
sociology. C. Wright Mills’ Sociological Imagination was still a daring
manifesto, his Power Elite still a model inquiry. Theory was predominantly (as
it was then called) ‘structural-functionalist’ and especially Parsonian. ‘Empir-
ical’ sociology was still predominantly ‘Columbia-oriented’ rather than
‘Chicago-oriented;’ data analysis was multivariate, not regression-based. Blau
and Duncan’s The American Occupational Structure was still half a decade to
a decade away. And social psychology was in large measure a choice between
‘small groups’ of the Bales variety or of the Michigan group dynamics variety,
a substantial dollop of ‘public opinion’ or ‘attitudes’ research, with a minority
voice somehow identified - often wrongly - with symbolic interactionism:
Blumer at Berkeley being the most visible - or vocal - representative, Goffman
(‘ The Presentation of Self in Everyday Life, published in the United States in
1959) just beginning to be recognized, Becker still largely unknown.

In anthropology, the Gumperz/Hymes special issue of the American
Anthropologist was not to appear until 1964, ethnoscience and componential
analysis were just coming into their own, the ethnography of communication
was just beginning to recruit its hoped-for army to canvas the world.

In linguistics, Chomsky’s Aspects of the Theory of Syntax was not published
until 1964, outsiders were just registering the import and impact of his review
of Skinner’s Verbal Behavior and his Syntactic Structures (1957). Linguistics
was just beginning to establish a track record for its significance to other
disciplines.

Throughout his stay in Berkeley, Sacks remained in touch with Harold
Garfinkel (now returned to his home base at UCLA) whose program of
ethnomethodological studies was being developed in a series of writings which
were privately circulated for the most part in mimeographed form. (It should
be recalled that it was not until 1959 that Garfinkel’ s ‘Aspects of the problem
of commonsense knowledge of social structure’ was published - and not in a
broadly accessible outlet at that; not until I960 for ‘The rational properties of
scientific and commonsense activities,’ also not in a source generally read by

Introduction

xv

sociologists; not until 1964 that ‘Studies of the routine grounds of everyday
activities’ appeared in Social Problems-, and not until 1967 that Studies in
Etbnomethodology was published.) It was largely through Sacks that these
manuscripts came to be circulated in Berkeley, largely among graduate
students in sociology. Of course, Sacks did not only circulate Garfinkel’s
manuscripts; in discussions among the students he added the special directions
of his own thinking, in some respects converging with Garfinkel’s, in other
respects quite distinctive.

At the time, Garfinkel was co-principal investigator with Edward Rose of
the University of Colorado on a research grant which supported a series of
conferences in Los Angeles in which Sacks took part. So Sacks’ engagement
with Garfinkel’s manuscripts in northern California was complemented by
more direct, personal engagement in the south. At the same time, other
developments were in progress in both north and south; in the north, for
example, Selznick had brought into his new Center for the Study of Law and
Society a number of graduate students in the social sciences, and especially
sociology. During the 1962-3 academic year, this group included Sacks,
David Sudnow and the present writer, whose activities separately and
together were to contribute to future developments, but are not directly in
point here.

In 1963, Garfinkel arranged for Sacks to move to Los Angeles. He was to
have an appointment as Acting Assistant Professor of Sociology at UCLA,
with the first year off. During that year, 1963-4, Garfinkel and Sacks 7 were
to serve as Fellows at the Center for the Scientific Study of Suicide in Los
Angeles, under the sponsorship of its director, Edwin Schneidemann. As it
happens, my own work prompted a move from Berkeley to Los Angeles
during the summer of 1963, and Sacks and I continued both a work and a
personal relationship during that year. I can therefore describe, at least in brief
compass, his primary intellectual preoccupations during the year. A great
many of them had his involvement with the Suicide Prevention Center as a
point of departure, thereafter taking the often surprising directions which his
distinctive mind imparted to them. In diverse ways, these interests show up
in his first ventures in teaching, the 1964-5 lectures which provided the point
of departure for the further development of the work, presented in the
subsequent lecture series published here.

One line of these concerns focussed on an examination of psychiatric, and
especially psychodynamic, theorizing, which furnished one primary theoreti-
cal handle on the phenomenon of suicide at the SPC, and which, more
particularly, was key to the so-called ‘psychological autopsies’ which were
conducted following suicides and which were of very great interest to both
Sacks and Garfinkel. Thinking about psychodynamic theorizing led Sacks (as
it had led me; cf. Berkeley Journal of Sociology, 1963) to a concern with
dialogue, and in particular with Platonic dialogue as a form of discourse
designed to control conduct. That, in turn, led him to a more general interest

7 And Erving Goffman, visiting on an occasional basis from Berkeley.

XVI

Introduction

in Greek philosophy, and particularly in Greek logic (on which he was
reading, among other sources, Kneale and Kneale’s The Development of Logic,
1962).

From the Freudian theorizing, from a prior interest in ‘children’s cultures’
set off by the work of the Opies (1959), and from a persistent attention to the
problems posed by the apparent facts and achievements of socialization, there
developed an interest in the behavior of children. This interest Sacks pursued
largely through examination of source books on children’s games (an interest
prompted as well by the work of O. K. Moore on games as ‘autotelic folk
models;’ cf. Anderson and Moore, I960), of the studies and protocols of
Barker and Wright (1951, 1954), in observational studies Sacks acquired
from Roger Brown, and other sources.

And Sacks pursued a number of other scholarly interests, in biblical studies
and interpretation, in translation, in archaeology, etc. In a very different vein,
Sacks came across stenographic transcripts, and then the tapes, of the
telephone calls to the Suicide Prevention Center of, or about, suicidal persons.
All of these themes may be found in the 1964-5 lectures, but it was the last
of them which provided the proximate source for the focussed attention on
talk itself - perhaps the most critical step toward the development of
conversation analysis.

Throughout the 1963-4 academic year, Sacks and I continued the
discussions and explorations entered into in Berkeley during the preceding
year and a half. This is not the place for a substantial account of those
activities (on-site explorations of the possibilities of field observation at the Los
Angeles International Airport, in the reference room of the UCLA library, at
neighborhood ‘Okie’ bars in Venice, and elsewhere; long discussions on the
UCLA campus where I was a visiting scholar, at the beach in Venice where
he lived, or at the apartment at the fringe of Beverly Hills where my wife and
I lived). But it may be of interest to describe what seemed to me at the time
something quite new, and seems to me now in retrospect the first appearance
of what would eventually become, after a number of major transformations,
what is now called ‘conversation analysis.’

It was during a long talking walk in the late winter of 1964 that Sacks
mentioned to me a ‘wild’ possibility that had occurred to him. He had
previously told me about a recurrent and much discussed practical problem
faced by those who answered phone calls to the Suicide Prevention Center by
suicidal persons or about them - the problem of getting the callers to give
their names. Now he told me about one call he had seen/heard which began
something like this:

A: This is Mr Smith, may I help you.

B: I can’t hear you.

A: This is Mr Smith .

B: Smith.

After which Mr Smith goes on, without getting the caller’s name. And later,

Introduction xvii

when Mr Smith asks for the caller’s name, the caller resists giving it. On the
one hand, Sacks noted, it appears that if the name is not forthcoming at the
start it may prove problematic to get. On the other hand, overt requests for
it may be resisted. Then he remarked: Is it possible that the caller’s declared
problem in hearing is a methodical way of avoiding giving one’s name in
response to the other’s having done so? Could talk be organized at that level
of detail? And in so designed a manner?

A month or two later, I arrived home at our apartment in the late
afternoon, to find Sacks waiting for me there. A transient difficulty with
Garfinkel had led him to realize that, if not on the present occasion then at
some future time, he might have to fend for himself in the academic
marketplace and had better have some written work to show. So he had
drafted the sketches of two papers. I left him talking to my wife in the living
room and retreated to my study and read the sketches. One of them was
about a methodical way of avoiding giving one’s name. As the reader who
turns to the 1964-5 lectures will soon discover, this is where Sacks’ lectures
began (not only in the composite version assembled for this publication, but
in the original as well).

Why might this episode, and these observations, be treated as the
beginning of what would come to be called ‘conversation analysis’? 8 Because

8 In the ‘General Introduction’ lecture for Fall 1967, (p. 621), Sacks introduces
the work to be presented by describing “When I started to do research in
sociology ...” It is unclear what Sacks means to refer to: when he went to
Cambridge? to Berkeley? sometime during graduate school? in Los Angeles? are these
the right terms to locate the reference?

In a way, the 1963 paper ‘Sociological description’ is not incompatible with the
account offered in this Fall 1967 lecture, except for the description (p. 622) of
starting “to play around with tape recorded conversations,” which surely did not
happen until the year at the Suicide Prevention Center. Until then, friends of Sacks
will remember occasions of sitting ‘with him’ in some public place and suddenly
realizing that Sacks was no longer in the same interaction, but was overhearing a
nearby conversation, and often taking out the omnipresent little multi-ring notebook
and jotting down a fragment of the talk and some observations about it. The virtues
of “replayling] them . . . type[ing] them out somewhat, and studying] them
extendedly” (Fall 1967, ibid.) were realized against a long experience of such
overhearing and notetaking. (One shared experience which may have alerted Sacks to
the payoffs of taking materials like the SPC tapes seriously was my experience during
1962-3 in Berkeley at the Law and Society Center of tape recording psychiatric
competency and criminal insanity examinations for subsequent analysis.)

But it is worth noting that Sacks did not set out to study conversation or language
in particular. His concern was with how ordinary activities get done methodically and
reproducibly, and the organization of commonsense theorizing and conduct which
was relevant to those enterprises. Clearly, he found talk, or what was being done
through talk, of interest before coming upon taped materials - else he would not
have been jotting overheard bits in notebooks. But the taped material had clear
attractions when it became available as a resource, and the talk invited being dealt
with as an activity in its own right. But that was something that turned out from
experience, not something that had been aimed at, or ‘theoretically projected.’

xviii Introduction

there is the distinctive and utterly critical recognition here that the talk can be
examined as an object in its own right, and not merely as a screen on which
are projected other processes, whether Balesian system problems or Schutzian
interpretive strategies, or Garfinkelian commonsense methods. The talk itself
was the action, and previously unsuspected details were critical resources in
what was getting done in and by the talk; and all this in naturally occurring
events, in no way manipulated to allow the study of them. And it seemed
possible to give quite well-defined, quite precise accounts of how what was
getting done was getting done - methodical accounts of action.

This was just the start of a long train of quite new things that Sacks was
to provide. It was only a little over a year later that the eventually published
version of ‘An initial investigation . . .’ (1972a) was completed. It is hard
now to appreciate how startlingly new and unprecedented that paper was at
the time. If one recalls the publication history of Garfinkel’s work (and that
GofFman’s Behavior in Public Places was published in 1963, and Relations in
Public was not to be published until 1971), a sense of its uniqueness when it
was published in 1972 might be somewhat more accessible. Its utter
originality in 1964-5 when it was being written, and the originality of the
materials in the first of these lectures which were delivered around the same
time, may be better grasped by reference to this other work. With the current
wisdom of hindsight, of course, our sense of this early work of Sacks’ is readily
assimilated to the direction we now know such studies took. But the
originality was not only startling in 1964 and 1965; it had the additional
headiness — and vertigo - of indeterminateness: How might one proceed?
What sort of discipline was this or might it be? Once a previous sense of
plausibility about the depth and detail of organization in conduct and
apperception of the world were set aside, what constraints on inquiry were
defensible? To what level of detail was it sensible to press?

During the summer of 1964, I left Los Angeles for the mid-west,
wondering what ever Sacks would do about lecturing to UCLA undergrad-
uates, and wondering as well how our contact could be sustained. The latter
problem was solved in part by a variety of resources that allowed me recurrent
trips to California during the 1964-5 year (though less so in ensuing years),
and in part by a practice which also satisfied my curiosity in the first respect.
Sacks would tape record his lectures and send them to me, and (if I remember
correctly) to David Sudnow who was spending the year in St. Louis, doing the
field work for his dissertation, later to appear as the book Passing On (1967).
At irregular intervals I would receive in the mail a little orange box with a
yellow label, containing a three-inch reel of tape, enough for the 50-minute
lectures (more or less) which Sacks was delivering. The lectures were for me,
then, a rather special form of monologic telephone call interspersed with our
dialogic ones (which were not recorded), and then, after Gail Jefferson started
transcribing the lectures, they were a sort of long letter series. 9 It turns out that

9 At the time they were being delivered, I encountered the lectures term by term,
like long analytical letters from Sacks. I had little overall view of them and of their

Introduction xix

they became Sacks’ most successful and prolific form of scientific communi-
cation.

When he wrote papers, Sacks imposed standards of formality and precision
that were extremely hard for him to meet to his own satisfaction. Most of the
papers he published under his own name alone were work-ups of lectures . 10
Most of the papers he drafted on his own as papers he was never sufficiently
satisfied with to publish. The exceptions, ‘An initial investigation . . . ’ or ‘On
some puns, with some intimations’ give some idea of what Sacks thought a
finished piece of work might look like.

Aside, then, from his collaborative publications, the lectures are the vehicle
by which most of his work was made available. Perhaps it was the explicitly
and necessarily informal and limited character of the occasion that could allow
him to get ‘the stuff’ out the best he could, with no pretense to finally
getting it ‘just right.’ Those who have seen some of his successive versions
of the ‘same pieces’ will know how great a change could overtake some
piece of work under the guise of getting it just right . 1 1 But the quality of
what was delivered in those lectures, and in those which followed, and the

overall development, of long term changes in the work reflected in them, etc. This
was largely because such changes would have come up in, or (without necessarily
being explicitly discussed) informed, our conversations with each other in the interim
between shipments, or could not be recognized for the changes they represented until
later developments. Largely, then, my reading was marked by my being struck,
charmed, and often amazed at what Sacks’ sleight of hand could materialize out of
a bit of data, the twist he could impart - no, discover - in it, the tacit understandings
he could, by a flash of insight, show we (‘casual’ readers or onlookers) had furnished
it. Sometimes the ‘twist’ assumed the proportions of a whole analytic topical area -
e.g., storytelling structure. I came to the reading of each new ‘package’ with a kind
of avid curiousity about what sorts of new things - whether unexpected observations
about a moment or whole new analytic issues - were tucked into those pages, and the
reading proceeded from flash to flash. It was like watching one’s athletic friend show
what he could do.

Preparation of this publication and this introduction has afforded me the occasion
for a larger overview, or series of overviews - of each set of lectures and of the set of
sets. In them I am brought to recall or to discover in retrospect larger scale
movements and changes, emerging and waning themes. Of course, this is refracted
through my own experience and intellectual colleagueship with Sacks. I have tried to
strike a balance between that kind of perspectival account and a less personalized
overview and setting-into-context.

10 Cf. for example, the paper on story-telling (1974). The paper on puns (1973)
is an exception here, having never been fully worked up as a lecture before being
prepared for the Georgetown Round Table, in whose proceedings it was published.
‘Everyone has to lie’ (1975) was adapted from a lecture, but the materials for the
lecture were initially drafted as a paper, under the title ‘The diagnosis of depression,’
which was never published in its original ‘paper’ format.

11 See the initial two lectures of the Spring 1966 term presented in this edition
with Sacks’ first effort at revision, at pp. 236-46 below, for a sample. In this case,
a virtually identical version of the same material was eventually published as ‘On the
analyzability of stories by children’ (Sacks, 1972b).

xx Introduction

special vision that underlies it, did not require getting it ‘just right’ to be
apparent. 12

Although he continued to tape a variety of research and teaching activities,
Sacks stopped recording his lectures in 1972 for a number of reasons. Some
of his lectures at the Linguistic Institute of 1973 at the University of Michigan
were recorded, as were some of the sessions of the joint seminar we taught, but
these were not recorded by Sacks, and were not reviewed for transcription by
him.

Harvey Sacks was killed in an automobile accident in November 1975
while on his way to the campus of the University of California, Irvine where
we were to meet to formulate a program which we were discussing
establishing at the Santa Barbara campus of the University. One can hardly
imagine what the next years of Sacks’ intellectual life would have produced,
especially in an academic environment fully supportive of the enterprise which
had already developed.

II

The ‘first installment’ of these lectures - the ones delivered during the
1964-5 academic year - can be furnished with two sorts of intellectual
reference points - ones in Sacks’ own intellectual development and ones in the
intellectual context around him.

In his own thinking, these lectures come after his paper ‘Sociological
description’ (1963), written in 1962-3 in Berkeley 13 and during the same
period as ‘An initial investigation . . (1972a) which was finished in June

1965. 14

Several features of these early papers which serve as landmarks in Sacks’
intellectual terrain, and of the early lectures, display some of the most potent
influences on his thinking at that time. There is first of all a wide-ranging
responsiveness to Garfinkel’s thematics, broadly acknowledged in a footnote
to ‘Sociological description’ (1963: 1), and in recurrent notes in the early
writings and lectures. A thorough treatment of the influences here, I daresay
the reciprocal influences at work here, remains to be written. At a different
level, there is the transparent allusion to the later Wittgenstein embodied in

12 Still, readers should bear in mind the in-progress status which this work had for
Sacks. While still alive, he expressed a willingness to have the lectures published, if
the publication could be done without much editing, not only because he did not
want to spend the time, but also to avoid masking the work-in-progress nature and
status of the effort. It should be a way of getting ‘a lot of stuff’ noticed, without
suggesting what should in the end be fashioned from it. The lectures were not meant
to look finished.

13 See the discussion below of the Spring 1966 lectures, and of ‘possible
description’ in particular.

14 Cf. the initial footnote to the version published in Sudnow (1972b).

Introduction

XXI

the invention (ibid.) of the ‘commentator machine’ as a grand metaphor for
(variously) the relationship of social science discourse to the social world which
is its object, of commonsense or lay talk about the world to ordinary
enactments of it, etc.

Perhaps less expectable in the contemporary academic setting, in which
studies of discourse and conversation are often set in contrast to transforma-
tional grammar, is the echo of generativist studies in the form of some of this
early work, and especially in the form of its problem development. Take as
a case in point ‘On the analyzability of stories by children’ (1972b, revised
from the first two lectures for Spring 1966, but first worked up as lectures 1
and 2 for Fall 1965).

The data for that set of lectures and publication, it will be recalled, were
taken not from ordinary conversation, but from the response of a young child
to a request by an adult for a story. Most relevantly for the present discussion,
this had the consequence that there was no ensuing talk by a co-participant
which could be examined to reveal an understanding of the ‘story’ which was
‘indigenous’ to the interaction, along the lines exploited in later conversation-
analytic work. In its place, Sacks relied on his understanding of the text being
examined (“The baby cried. The mommy picked it up”), and the under-
standing which he attributed to his audience - understandings not overtly
provided for by the text itself (for example, that ‘the mommy’ is the mommy
of that baby, although the story as told by the child was expressed as ‘ the
mommy picked it up’).

The problem, as Sacks developed it, was to build ‘an apparatus’ that would
provide for such hearings or understandings, and would serve both as a
constraint on them and as a research product to which they could lead. This
form of problematics, of course, echoes the commitment to build a syntactic
apparatus which would provide for the alternative parsings of a claimedly
ambiguous sentence such as ‘Flying planes can be dangerous’ (Chomsky,
1957). The reader is first asked to recognize that alternative ‘structural
interpretations’ can be assigned to this sentence, and then to be concerned with
the construction of a syntax that produces such an ambiguity and provides for
its disambiguation. To be sure, this form of problem development and
statement is invoked by Sacks on behalf of a quite different intellectual and
scientific enterprise, but the formal similarities in the problematics seem clear
enough. 15 (And connections appear in other guises as well, for example, in the

15 See the comments on the Fall 1965 lectures for further discussion of the
relationship to generative grammar studies.

In this regard as well, John Heritage has called to my attention an exchange
involving Sacks and others at the Purdue Symposium on Ethnomethodology, in
which he remarks in response to several inquiries (Hill and Crittendon, 1968: 41-2),

One of the things that is obvious from the kind of analysis I have given you
is that there can be a set of rules which can reproduce the problems in the data
with which you started . . . [Query: How do you become satisfied with a
solution?] ... I have a set of rules which give me back my data.

xxii Introduction

extensive paragraph numbering system which is used to organize ‘An initial
investigation . . as well as the ‘Introduction’ of 1965 printed in this
volume), a format hardly familiar to sociologists at all, but in common
practice in linguistics at the time, though Sacks may well have come to it
initially through his study of the law).

I think it characteristic of Sacks’ relationship to work which he respected
that it would enter into the warp and woof of his own thinking and would
shape the way he did his work. And this is so not only in this formative stage
of his work. Later on (in the work published in Schenkein (1978) but
delivered as lectures 9-12 in Fall 1971), for example, his argument that the
obscenity in a dirty joke is not its point, but is rather a form of ‘circulation
control’ on knowledge which is packed or tucked in elsewhere, not overtly
labelled or featured as the point of the joke, brings to bear a form of analysis
developed by scholars of classical Greece such as Milman Parry (1971) and
Eric Havelock (1963) in work on the role of the Homeric epics in an oral
culture and its transformation in the passage from an oral to a literate culture.

Another case in point is furnished by Cressey’s work on embezzlement
(1953), which served Sacks (in ‘An initial investigation . . . ,’ 1972a) not to
constitute the problem or suggest the shape of a solution, but as a way-station
in the substantive analysis. Cressey had proposed as a key to understanding
embezzlement that its perpetrators all had ‘a nonsharable problem.’ In Sacks’
effort to come to terms with the assertion by some avowedly suicidal persons
that they had ‘no one to turn to,’ he proposed as a proximate solution that
these persons found that what troubled them would, if recounted to the ones
they would properly turn to (e.g., spouse), undermine the very relationship
that made them ‘turnable-to;’ that is, precisely, they had a ‘nonsharable
problem.’ But for Sacks this merely served to pose a problem: how to
formulate the terms of the ‘search for help’ that yielded these persons as the
candidates to be turned to, and therefore yielded the result that a problem not
sharable with them left the searcher with ‘no one to turn to.’ And that
recasting of the problem led to the central contribution of that analytic
undertaking - the formulation of ‘membership categorization devices’ and
their features. 16

Sometimes Sacks would cite such sources. More often, the shape of the
problem formation or solution, or the analytic resource, had simply entered
into the currency of his thinking, and its source was lost sight of, especially in
the context of lectures to undergraduate courses. The lecture format is, in this
regard, ‘informal.’ Although published work which is, taken as a whole,
remote from his concerns is often quoted directly and/or cited by name (e.g.,
Freud, Gluckman, Von Senden), more intimately related work is often not,

16 In the paper presenting this work (Sacks, 1972a), the analytic ordering given
in the text here is reversed. The paper begins with the most formal and general posing
of the issues of categorization, and only eventually arrives at the more proximate,
situated problem, as a ‘derivation,’ i.e., the dilemma presented when what qualifies
another as the proper person to turn to will be compromised by the very turning to
them.

Introduction xxiii

as for example (to cite an early instance in the text which follows) in the
discussion of ‘common knowledge’ in lecture 3 of the 1964-5 lectures (as
printed herein), for which Garfinkel clearly was relevant. In the preparation
of these lectures for the present publication, that practice has not been
addressed; it is a characteristic feature of the form in which Sacks’ work was
shaped for presentation.

As unexpected as may be the appearance in Sacks’ early lectures of echoes
of the analytic style of transformational grammar, even more striking is the
apparent lack of specific influences from the work of Erving Goffman. This is
especially surprising since, during the years at Berkeley, Sacks took Goffman
more seriously than he did virtually any other member of the faculty.

At a very general level, of course, Goffman’s analytic enterprise had
undertaken to establish the study of face-to-face interaction as a domain of
inquiry in its own right, and his work was very likely central in recruiting
Sacks’ attention to face-to-face interaction as a focus for the concern with
practical theorizing and commonsense reasoning which animated the eth-
nomethodological enterprise. Surely Sacks’ work, and work which it in-
spired, have been important to whatever success and stability this area of
inquiry has achieved. And Sacks could treat Goffman ’s work as setting a
relevant domain for students for pedagogical purposes; in the first of the Fall
1967 lectures, Sacks recommends readings in Goffman’s work as the most
relevant sort of preparatory reading for the course, and the most indicative of
the general stance of the course, while explicitly differentiating his own work
from it.

Goffman’s influence on Sacks was at its peak during Sacks’ years as a
graduate student. While at Berkeley, for example, Sacks satisfied a require-
ment in one of Goffman’s courses not with an empirical study of interaction
of the sort chracteristic of his later work, but by writing the so-called ‘police
paper’ (later published as ‘Notes on police assessment of moral character,’
1972c), concerned with methods of commonsense theorizing about appear-
ances and moral character, and based on handbooks and manuals of police
procedure. The subsequently published version of the paper begins with a
handsome acknowledgement of debt to Goffman’s writing and lectures, and
though the style and ‘address’ of the work differ in various respects from those
of Goffman, the topic plays off of several themes recurrent in Goffman’s work
at the time, and the exploitation of handbooks and manuals echoes Goffman’s
use of manuals of etiquette and advice. But after this, Sacks’ work diverges
increasingly from Goffman’s.

To be sure, in later work Sacks addressed himself to more specific
interactional topics mentioned in Goffman’s work (see, for example, the
discussion of ‘rules of irrelevance’ in Goffman’s essay ‘Fun in games,’ (1961:
19ff), or the passing mention of turn-taking (Goffman, 1964: 136), but the
lines of influence are often not entirely clear. Goffman is reported to have
responded to a question years later asking whether Sacks had been his student
by saying, “What do you mean; I was his student!” Leaving aside the possible
elements of generosity, irony and flipness in such a remark (and assuming

XXIV

Introduction

that the report is, generally speaking, correct), a serious treatment of the
directions of influence and the interplay of ideas between them remains to
be written. 17

That important divergences between Goffman and Sacks began to develop
early on can hardly be doubted. These came to a head, both symbolic and
practical, over Sacks’ PhD dissertation, an episode which cannot be recounted
here. 18 For now the upshot must remain this: although in retrospect Sacks
seems clearly to have labored in the same vineyard, and although he was not
only formally Goffman’s student but learned a great deal from him, the
degree to which Goffman influenced more specifically the work for which
Sacks is known remains an open question. Certainly, such specific influences
are not as much in evidence as most readers are likely to expect, either with
respect to Goffman’s most characteristic substantive concerns - face, de-
meanor, structures of attention and information, etc., with respect to
governing themes - dramaturgic, ethologic, frame-analytic, etc., or with
respect to data and method.

Ill

In mentioning genres of work and particular people who constituted a
relevant intellectual ambience for the early corpus of Sacks’ work, one name
which might be thought missing is that of John Searle. But it turns out that
Searle’s work constitutes a parallel stream, not a source. Indeed, although his
Speech Acts was published in 1969, his paper ‘What is a speech act?’ appeared
in 1965, the same year as the first of Sacks’ lectures. It is striking to compare
the quite different tacks taken in these two approaches to the accomplishment
of social action through the use of language, even if only in the brief and
superficial way that space limitations compel.

Searle begins not with a particular utterance - either actually spoken or
invented. He addresses himself rather to a class of utterances that would
satisfy whatever is required for them to effectively - felicitously - accomplish
the speech act of ‘promising.’ It is the type ‘promises’ that provides Searle his
object of inquiry. The solution takes the form of stating the “conditions . . .

17 Some considerations on the relationship between Goffman’s work and
conversation analysis may be found in Schegloff (1988). Goffman’s most explicit
engagement with conversation-analytic work appeared in Forms of Talk (1981), the
earliest of whose essays dates to 1974.

The upshot was that Goffman found the argument of ‘An initial investiga-
tion . . .’ circular, and no amount of discussion could move him from this view. Nor
would he, for quite a while, step aside from the committee to allow its other members
to act favorably on the dissertation, as they wished to do. Eventually, however, he
agreed to do so, largely at the urging of Aaron Cicourel who, in the end, signed the
dissertation as Chair of its sponsoring committee, making possible the awarding of
the PhD in 1966.

Introduction

xxv

necessary and sufficient for the act of promising to have been performed in the
utterance of a given sentence” (i.e., a general definition of ‘promise’), with a
later derivation of the rules for performing acts of this class.

Readers may recall the sort of result yielded by proceeding in this manner
- the formulation of preparatory conditions, sincerity conditions, etc.,
followed by ‘‘rules for the function indicating device for promising.” The
focus, then, is on the class or type of act, and the term describing it -
‘promising.’ It is not on particular utterances or the contexts in which they
occur. Indeed, Searle’s paper begins by invoking the most general context
possible, ‘‘In a typical speech situation involving a speaker, a hearer, and an
utterance by the speaker ...”

Sacks’ first lecture starts in a significantly different way (and although the
original transcripts show a much more uneven presentation than appears in
the edited version, in the manner of their opening they do not differ). Sacks
begins by offering particular utterances in a particular context. Our attention
is focussed from the outset on particular exchanges, such as A: “Hello,”
B: “Hello;” or A: “This is Mr Smith, may I help you;” B: “Yes this is
Mr Brown;” or A: “This is Mr Smith, can I help you;” B\ “I can’t hear you,”
which are

. . . some first exchanges in telephone conversations collected at an
emergency psychiatric hospital. They are occurring between persons
who haven’t talked to each other before. One of them, A, is a staff
member of this psychiatric hospital . . .

Sacks goes on to offer a variety of detailed considerations about what these
utterances, “This is Mr Smith,” “can I help you” or “I can’t hear you” might
be observed to be doing, and how they might be doing it. Then he remarks
(lecture 1, pp. 10-11):

Clearly enough, things like “This is Mr Smith,” “May I help you”? and
“I can’t hear you” are social objects. And if you begin to look at what
they do, you can see that they, and things like them, provide the
makings of activities. You assemble activities by using these things. And
now when you, or I, or sociologists, watching people do things, engage
in trying to find out what they do and how they do it, one fix which can
be used is: Of the enormous range of activities that people do, all of
them are done with something. Someone says “This is Mr Smith” and
the other supplies his own name. Someone says “May I help you” and
the other states his business. Someone says “Huh?” or “What did you
say?” or “I can’t hear you,” and then the thing said before gets repeated.
What we want then to find out is, can we first of all construct the objects
that get used to make up ranges of activities, and then see how it is those
objects do get used.

Some of these objects { recall that ‘objects’ here refers to the utterances
which have been examined } can be used for whole ranges of activities,

xxvi Introduction

where for different ones a variety of the properties of those objects will
get employed. And we begin to see alternative properties of those
objects. That’s one way we can go about beginning to collect the
alternative methods that persons use in going about doing whatever
they have to do. And we can see that these methods will be reproducible
descriptions in the sense that any scientific description might be, such
that the natural occurrences that we’re describing can yield abstract or
general phenomena which need not rely on statistical observability for
their abstractness or generality. 19

Nor (one might add) do they rely for their abstractness or generality on being
stripped of all contextual particulars (in the manner of Searle’s “In the typical
speech situation . . .”) or on the stipulation of general constitutive definitions
of verbs for speaking.

The focus in Sacks’ work here, and in much of the work of the ensuing
years, 20 is not on general constitutive conditions, or even on rules in Searle’s
sense, but on practices and methods - on how Members, in particular contexts
(or classes of context arrived at by examining particular contexts), methodi-
cally construct their talk so as to produce a possible instance of an action or
activity of some sort, and to provide for the possible occurrence next of
various sorts of actions by others.

Although the 1964-5 lectures exhibit some striking early explorations
along these lines, a particularly exemplary instance of such an analysis is
Sacks’ discussion in lecture 4 of Spring 1966, of the utterance by a previously
present participant, after a newcomer to a group therapy session of teenaged
boys has been greeted, “We were in an automobile discussion,” which Sacks
undertakes to show to be “a possible invitation.” (In later ‘takes’ of this
analysis, the treatment is varied; for example, in Fall 1968, lecture 6, (volume
II) he discusses it as ‘orientational,’ although all the analysis bearing on its
‘invitational’ aspect is included. This later discussion is rather fuller, more
detailed and compelling.)

His undertaking — . to build a method which will provide for some

utterance as a recognizable invitation ...” - may sound like Searle’s, but it
turns out to be quite different. There are two component tasks. One of these
tasks is

19 The reference to “reproducible descriptions in the sense that any scientific
description might be” is an appearance in this first lecture of a theme and argument
which Sacks had been percolating for some time, and which was written up at the end
of the 1964-5 academic year in a putative introduction to a publication which never
materialized. (That introduction is included in this volume, and its argument is
recounted below, at pp. xxx-xxxii.)

20 When Sacks does introduce a shift to a rather more general form of
undertaking, for example at lecture 3 of the Fall 1968 set, it still has quite a different
character than Searle’s.

Introduction

xxvu

... to construct ... “a partial definition of an invitation.” What makes
it partial is that while it’s a way of doing invitations, it’s not ... all the
ways . . . there are other ways and those would be other partial
definitions.

The second task is to have this partial definition provide for the actual case
which occasions the inquiry:

We want to do both: Construct a partial definition of ‘invitation,’ and
one that provides for ‘this is a case.’

It turns out that there are other things such an analysis should do, which need
not preoccupy the present discussion.

The construction of the method that provides for the data under
examination as a possible instance of ‘invitation’ has two parts. First, Sacks
characterizes the ‘slot’ in which this utterance occurs, and characterizes it in
various ways - as (1) just after introductions and greetings, (2) in the arrival
of a newcomer to a conversation already in progress, (3) in a situation of a
psychiatric neophyte coming to group therapy for the first time and joining
more experienced patients, etc. Second, he characterizes one particular aspect
of the utterance itself - its formulation of the topic preceding the newcomer’s
arrival as “an automobile discussion.” He shows that that formulation makes
relevant the common category membership of the newcomer and the others,
but a category membership as “teenaged boys” or potential “hotrodders,”
rather than as “patients.” And in formulating the topic as one for which the
newcomer might be competent in common with them (rather than as one for
which he is not, as is done by a next speaker who extends the utterance by
saying “. . . discussing the psychological motives for . . .”), a possible invita-
tion is done.

What this (here highly oversimplified) analysis provides, then, is not
necessary and sufficient conditions for the felicitous performance of an
invitation, or rules for its performance, but rather a partial method (Sacks
refers to it as a “a partial definition”) for doing an invitation in a particular
interactional / sequential context.

IV

As noted, the earliest lectures, of 1964-5, include a variety of efforts to
develop analyses along these lines. Certain themes recur, only some of which
can be remarked on here, to highlight something of an abbreviated catalogue
of concerns animating Sacks’ work at the time.

Consider, for example, the following sort of issue to which Sacks addresses
himself recurrently throughout the 1964—5 lectures (this is not an exhaustive
listing):

How to get someone’s name without asking for it (give yours), lecture 1.

xxviii Introduction

How to avoid giving your name without refusing to give it (initiate repair),
lecture 1.

How to avoid giving help without refusing it (treat the circumstance as a
joke), lecture 2.

How to get an account without asking for it (offer some member of a class
and get a correction), lecture 3.

How to get people to show they care about you, given few opportunities
afforded by routine life, e.g., of the divorced (commit/attempt suicide),
lecture 5.

How to introduce a piece of information and test its acceptability without
saying it, lecture 6.

How to do a ‘safe’ compliment, i.e., without derogating others, lecture 8.
How to get help for suicidalness without requesting it (ask ‘how does this
organization work?’), lecture 10.

How to talk in a therapy session without revealing yourself (joke), lecture

12 .

Sacks’ analytic strategy here is not a search for recipes, or rules, or
definitions of types of actions. He begins by taking note of an interactional
effect actually achieved in a singular, real episode of interaction (in the listing
above, this often includes an achieved absence - something which did not
happen). And he asks, was this outcome accomplished methodically. Can we
describe it as the product of a method of conduct, a situated method of
conduct, such that we can find other exercises or enactments of that method
or practice, in that situation or context or in others, which will yield the
accomplishment, the recognizable accomplishment (recognizable both to
co-participants and to professional analysts) of the same outcome — the same
recognizable action or activity or effect.

So in the listing I have offered above, the ‘solutions’ mentioned in
parentheses after some of the ‘problems’ are not ‘general;’ they are not
practices which whenever or wherever enacted will yield those activities as
systematic products. They are situated, contexted. How to describe the
relevant contexts, the scope within which the proposed practice ‘works’? That,
of course, is one of the prime sets of problems in this analytic enterprise. How
shall we as analysts describe the terms in which participants analyze and
understand, from moment to moment, the contexted character of their lives,
their current and prospective circumstances, the present moment - how to do
this when the very terms of that understanding can be transformed by a next
bit of conduct by one of the participants (for example, a next action can recast
what has preceded as ‘having been leading up to this’). Clearly enough, these
questions are of a radically different character than those which are brought to
prominence in an undertaking like that of Searle, or Austin (1962) before
him.

The recurrent theme documented above will remind some readers of
‘indirect speech acts.’ In many items on that list the problem appears to be
how to achieve some result without doing it ‘directly’ (as one says in the

Introduction

XXIX

vernacular - and it is a vernacular term). The proposed ‘solutions’ might then
be cast, in this vernacular and quasi-technical, idiom, as ‘indirect’ speech acts,
although this is, of course, an idiom not employed within the conversation-
analytic tradition, (cf. Levinson, 1983: 356-64 for one account).

One line of inquiry (ibid., 274; Brown and Levinson, 1978; Lakoff, 1973)
relates the use of indirect speech acts to considerations of politeness. But
Sacks’ discussion focusses instead on what might be termed ‘strategic/
sequential’ considerations. He notes that the sorts of next turns made relevant
by what might be called direct requests are quite different from the ones made
relevant by the conduct whose methodic practices he is explicating. When
answerers of the telephone at a psychiatric emergency service ask “What is
your name?” they may get in return a request for an account - “Why?” - and
may end up not getting the name. When they give their own names, they do
not get asked “Why?,” because they have not done an action which is
accountable in that way. The thrust of the analysis is, then, not considerations
of politeness, but contingent courses of action as progressively and differen-
tially realized in the set of turns that make up structured sequences based on
what would later come to be called ‘adjacency pairs.’ 21

The divergence of these two paths of analysis seems quite clearly related to
the materials being addressed. On the one hand, we have single classes of
utterances, and eventually (Searle, 1976) not even particular ones necessarily,
but the categorical type of action which they are supposed to instantiate,
singly and across contexts. On the other hand, we have particular utterances
occurring in particular series of utterances, in particular organizational,
interactional and sequential contexts, with the source of the utterance in prior
talk and conduct accessible and demonstrably relevant to professional/
academic analysis as it was to the participants in situ and in vivo , and with the
ensuing interactional trajectory which was engendered by the utterance
inviting examination in the light of the set of possibilities from which it might
have been selected. One of these sets of materials is the natural setting for the
work of philosphy and ‘academic’ inquiry; the other is rather closer to the
natural setting for the workings of talk in the everyday world. Sacks’ first
lectures make clear what course is being set.

The consequentiality of working with particular data, for example, with
particular utterances, is underscored elsewhere in these and subsequent
lectures, when Sacks directs the problematic of describing a ‘method for the
production of . . . ’ to whatever action label one would assign to an utterance
such as “I’m nothing.” Sacks asks (lecture 9, p. 67): how does someone
“properly and reproducibly” come to say such a thing, this thing? What is
someone doing by saying this thing, and how do they come to be doing it?

At the time that Sacks was launching inquiry along these lines, a common
reaction was that an utterance of this sort was ‘just a manner of speaking.’

21 Sacks deals with these themes from a different stance subsequently in Winter,
1967, cf. the lecture for March 9 in particular, and the discussion of the varying tacks
he takes below at pp. 1-li of this introduction.

XXX

Introduction

That the particular way of speaking, the phrasing, was almost accidental (a
stance suitable to the view that an utterance is an enactment of a sentence
which expresses a proposition, where it is the underlying proposition -
perhaps accompanied by its ‘function indicating device’ - which finally
matters, not the particulars which happen to give it expression on any given
occasion). But Sacks saw it as the outcome of a procedure, as announcing ‘a
finding’ by its speaker. He asked what that procedure was, and how it could
arrive at such a finding, in a fashion that other participants would find
understandable, and even ‘correct.’ He took seriously the particular form in
which conduct appeared - the participants had said this thing , in this way ,
and not in some other way. He insisted on the possibility that that mattered
- that every particular might matter. None could be dismissed a priori as
merely (a word he particularly treated with suspicion) a way of talking.

Of course, the fullest version of this sort of analytic undertaking was Sacks’
paper ‘An initial investigation . . .’ (1972a), where the utterance/action in
question was ‘I have no one to turn to.” This utterance was also seen as
reporting the result of a search, the description of which required developing
the terms in which such a search might be understood to have been
conducted, namely, ‘membership categorization devices.’ Early versions of
parts of that paper (as well as other papers) can be found in the 1964-5
lectures, for example in lecture 6.

This way of working, then, mixed a kind of naturalism (in its insistence on
noticing and crediting the potential seriousness of particulars of the natural
occurrences of conduct) with the ethnomethodological concern for the
Members’ methods for the production of a mundane world and commonsense
understandings of it. Sacks asked how the recognizably detailed ordinary
world of activities gets produced, and produced recognizably. It was just this
way of proceeding - describing procedurally the production of courses of
action - that Sacks understood at the time to be the foundation of the sciences
as ‘science,’ and therefore the grounds for optimism about the principled
possibility of a natural observational discipline in sociology. A brief account of
this view (argued in the ‘Introduction’ by Sacks, Appendix I in this volume)
is in order.

V

Sacks had developed an argument 22 addressed to the question of whether

22 The argument was written up, probably in the summer or autumn of 1965,
after Sacks’ first academic year of lectures, as a possible introduction to a
contemplated volume entitled The Search for Help. This publication, which was never
pursued, would have included two papers - The search for help: no one to turn to’
(later published in Sudnow, 1972b), and ‘The search for help: the diagnosis of
depression,’ never published. That the argument informed his thinking earlier, and
entered into the first lectures, can be seen in the excerpt from lecture 1 cited at
pp. xxv-xxvi above, and remarked on in n. 19.

Introduction

XXXI

sociology could be shown to be a possibly ‘stable’ natural observational
discipline. By this question Sacks meant to address the possibility that social
science provided merely stopgap accounts of human action, conduct, behav-
ior, organization, etc., until such disciplines as biology and neurophysiology
matured to the point at which they could deal with such problems. (This was
a position that Sacks was trying out when I first met him in 1961-2, and
could be seen as a kind of riposte to Chomsky’s critique of Skinner. I always
suspected that Sacks entertained the position as a provocation, in a law school
pedagogical way, rather than as seriously tenable, but used it to force a
consideration of the arguments necessary to set it aside. The position certainly
shook me up when Sacks first confronted me with it in the winter and spring
of 1962, for, in common with most sociology graduate students, I had treated
such claims as long since undermined by Durkheim and other ancestors.) If
sociology, or social science, were such a stopgap and thus ‘unstable,’ it hardly
seemed worth investing much time and commitment in it. So before setting
off on a serious research undertaking, it seemed in point to establish that a
stable discipline was possible. Sacks believed that the argument he developed
had a further pay-off; it showed something of the features the research
enterprise and its results should have if it were to be, or contribute to, a stable
science. The argument, briefly stated, was this.

Contributions to science, including to sciences such as biology and
neurophysiology, are composed of two essential parts. One is the account of
the findings. The other is the account of the scientists’ actions by which the
findings were obtained. What discriminates science from other epistemic
undertakings is the claim that its findings are reproducible, and that
reproducibility is itself grounded in the claim that the results were arrived at
by courses of action reproducible by anyone in principle. Other investigators
can, by engaging in the same actions, arrive at the same findings.

Sacks argued that both of these parts of contributions to science are
‘science’, and not just the findings. For it is the reproduction of the actions
reproducing the results which make the findings ‘scientific’, and the descrip-
tions of those courses of action which make their reproducibility possible. If
the results are scientific, the descriptions of the actions for producing them
must also be science.

But, he noted, the descriptions of courses of action in scientific papers are
not couched in neurophysiological terms, but take the form of accounts of
methods or procedures. This form of account of action is reproducible, both
in action and in description.

So, Sacks concluded, from the fact of the existence of natural science there
is evidence that it is possible to have (1) accounts of human courses of action,
(2) which are not neurophysiological, biological, etc., (3) which are repro-
ducible and hence scientifically adequate, (4) the latter two features amount-
ing to the finding that they may be stable, and (5) a way (perhaps the way)
to have such stable accounts of human behavior is by producing accounts of
the methods and procedures for producing it. The grounding for the
possibility of a stable social-scientific account of human behavior of a

xxxii Introduction

non-reductionist sort was at least as deep as the grounding of the natural
sciences. Perhaps that is deep enough.

This conclusion converges, of course, with the thrust of ethnomethodology
as Garfinkel had been developing it, and was undoubtedly motivated, at least
in part, by Sacks’ engagement with Garfinkel (and informed, perhaps, by
Felix Kaufmann (1944) as well). Still, the argument is novel and provides a
grounding from a different direction than Garfinkel had provided. For the
tenor at least of Garfinkel’s arguments was anti-positivist and ‘anti-scientific’
in impulse, whereas Sacks sought to ground the undertaking in which he was
engaging in the very fact of the existence of science. (And, indeed, in the
earlier ‘Sociological description’ (Sacks, 1963) he had written, “I take it that
at least some sociologists seek to make a science of the discipline; this is a
concern I share, and it is only from the perspective of such a concern that the
ensuing discussion seems appropriate.”)

VI

I have remarked on two types of problems taken up in the 1964-5 lectures
- the reproducible methods by which ‘findings’ such as “I’m nothing” or “I
have no one to turn to” may be arrived at (note in this regard the special claim
on Sacks’ attention exerted by commonsense uses of ‘quantifiers,’ starting
with the ones mentioned above,’ but extending to utterances such as ‘Everyone
has to lie’, (Sacks, 1975)), and how to achieve some outcome without aiming
for it ‘directly’. Several other recurrent themes in these earliest lectures might
be mentioned here.

One is an attention to certain ‘generic forms’ of statement or question, into
which particular values can be plugged in particular circumstances. Sacks
isolates, for example, the question form ‘Why do you want to do X?’ (lecture
5, p. 33), or the generic form of statement ‘Because A did X, B did Y’
(lecture 5, p. 36). Later he focusses on the form, ‘X told me to call/do Y’
(lecture 10, pp. 76-7). It was very likely the exposure almost exclusively to
calls to the Suicide Prevention Center, and the sort of recurrencies which they
provided, which led to a focus on regularities so literally formulated. But it
was in this sort of problem that the concern with the formats of utterances,
often rather more abstractly and formally described, initially appeared.

There is throughout these lectures the repeated use of ‘the socialization
problem’ as a resource for focussing analysis. The question gets posed, ‘How
does a child learn that X?,’ for example, that activities are observable; what
properties of competence does socialization have to produce, and how are they
produced; how does this learning take place (e.g., lecture 14, pp. 120-1).
This form of problem or observation finds expression in Sacks’ writing of this
period as well as in the lectures (for example, in the remarks in ‘An initial
investigation. . . .’ concerning what is involved in learning how adequate
reference is to be done), although it recedes in prominence in the later years
of the lectures.

Introduction xxxiii

These early lectures of 1964—5 touch on, or give a first formulation of, a
variety of themes more fully developed in later work, either of Sacks’ or by
others.

For example, although many believe that the early lectures were taken up
with membership categorization, and that sequential organization is only
addressed in later years, we have already seen that the early lectures -
including the very first - engage that issue from the very beginning. To cite
but one other instance of this early engagement, lecture 9 includes observa-
tions on sequence organization (the asker of a question gets the right to do
more talk), on what were later (Sacks et al., 1974) called contrasting speech
exchange systems (remarks on press conferences and cross-examination), on
how the turn-taking systems of different speech exchange systems can affect
the forms of utterances (e.g., long questions when there is no right of follow
up), and the like.

Or note how the earlier-mentioned recurrent theme concerned with ‘how to
do X without doing Y’ finds later resonance not only in Sacks’ work but in
work such as that by Pomerantz (1980) on ‘telling my side as a fishing device’
(how to elicit information without asking for it), by Jefferson (1983) on
‘embedded correction’ (how to induce adoption of a correct form without
correcting the wrong one), and others.

Or consider the material in lecture 1 1 concerned with glancing, looking,
and seeing. The parts of this discussion which concerned the categories in
terms of which one sees, anticipate the later discussion of ‘viewer’s maxims’
in the lectures on “The baby cried” (lectures 1 and 2 for Spring, 1966,
eventually published as ‘On the analyzability of stories by children,’ 1972b).
They display as well Sacks’ reflections on what such glance exchanges reveal
about ‘norms’ in the more conventional sociological and anthropological
sense, about ‘social integration,’ ‘alienation,’ and the like. And perhaps there
is here as well a point of departure for Sudnow’s later (1972a) work on
glances, for example in Sacks’ observation (p. 86) that “We start out with the
fact that glances are actions.”

It is worth noting that in some cases, discussions in these early lectures
include points that are not found in later elaborations. Some of these seem to
me to have been simply wrong - for example, the claim (lecture 5, p. 33) that
‘opinion’ is something you don’t need a defense for. Others encountered
problematic evidence within the conversation - analytic tradition of work. For
example, Sacks had proposed that a method for doing greetings consisted in
the use of one of the class of greeting terms in ‘first position.’ Schegloff (1967)
disputes the generality of the claim by examining telephone conversation in
which “Hello” in first turn is ordinarily not a greeting, and shows that claims
in this domain of work can be addressed with data, investigated empirically
and found to be the case or not.

Still other portions of these early lectures, however, appear to be strong
points which simply dropped out of later reworkings of the topic. For
example, lecture 6 is a version of (or draws on) ‘An initial investigation . . . ,’
‘On the analyzability . . .’ ‘Everyone has to lie,’ and a paper which Sacks

XXXIV

Introduction

never published, entitled ‘A device basic to social interaction,’ concerned with
the character of the categories which compose membership categorization
devices as organizing devices for commonsense knowledge about members.
But Sacks makes a point in this lecture which I do not believe ever appears
in any of the other accounts of these domains, concerning the relativity of
category collections such as age and class to the categorizer; he notes that the
recipient of some utterance which includes some such categories (such as
‘young man’) has to categorize the categorizer to know how they would
categorize the one who had been categorized in the utterance.

These lectures, then, have more than merely historical interest as embryonic
versions of later developed work. Some of the themes here, however insightful
and innovative, happen not to have been further developed. And others,
which were further developed, left behind some points which are still valuable
and can be found here.

VII

As with the 1964-5 lectures, those for the Fall 1965 term include first tries
at topics (both accounts of specific data episodes and analytical topics raised
from them) taken up and elaborated in subsequent terms, as well as
discussions which do not get such subsequent development. Among the latter
are, for example, ‘hotrodding as a test’ (lecture 10) or ‘non-translatable
categories’ (lecture 12). Among the former are “The baby cried . . and
membership categorization devices (lectures 1 and 2) more fully elaborated as
lectures 1 and 2 in the following term, Spring 1966; collaborative utterances
addressed via “We were in an automobile discussion” in Spring 1966
(lectures 4 and 5); or ‘tying rules’ taken up in a number of subsequent lecture
sets.

Still, there is good reason to read carefully the discussions of Fall 1965,
even for topics which are given fuller, and apparently more satisfactory,
treatment later. To cite but a single example, in the outline for the initial
lecture on “The baby cried ...” (here appearing as Appendix A for Fall
1965), at l.a.2 and a. 3, Sacks offers observations which do not appear in
subsequent treatments of this material (either in Spring 1966, or in the
subsequent publication as Sacks, 1972b) but which differentiate Sacks’ point
here from other, parallel claims - often characterized as being concerned with
the order of narrativity. Others (often more or less contemporaneously, e.g.,
Labov and Waletsky, 1966) have remarked that in narratives the ‘default’
organization is that order of sentences is isomorphic with the order of the
occurrences which they report. And in later versions of this analysis Sacks
seems to be making the same argument. As it is put in the published version
(1972b: 330), “I take it we hear that as Sfentence] 2 follows Sentence} 1, so
Occurrence] 2 follows Ofccurrence] 1.” But here, in the Fall 1965 outline,
he notes that “this cannot be accounted for simply by the fact that SI precedes
S2,” for “we can find elsewhere two sentences linked as these are, with

Introduction

XXXV

nothing between, where we would not hear such an action sequence.” And he
offers an instance from the same collection of children’s stories, . The
piggie got hit by the choo-choo. He got a little hurt. He broke his neck. He
broke his chin.”

The point is that what is at work here is more than a matter of narrative
technique or of discourse organization, although these may well be involved.
Rather ‘commonsense knowledge’ of the world, of the culture, and of
normative courses of action enter centrally into discriminating those actions or
events whose description in successive sentences is to be understood as
temporal succession from those which are not. It is not, then, a merely formal
or discursive skill, but can turn on the particulars of what is being reported.
This theme drops out of later discussions of these materials. 23

If this point seems to resonate basic themes of so-called contextualist, or
social constructionist or ethnomethodological stances, there are other elements
in these early lectures which operate on a different wavelength. For example,
early in the development of what he called ‘tying rules’ (in which he is
addressing matters later often discussed under the rubric of ‘cohesion,’ cf., for
example, Halliday and Hasan, 1976) he proposes (Fall 1965, lecture 5,
p. 159) to be

taking small parts of a thing and building out from them, because small
parts can be identified and worked on without regard to the larger thing
they’re part of. And they can work in a variety of larger parts than the
one they happen to be working in. I don’t do that just as a matter of
simplicity . . . the image I have is of this machinery, where you would
have some standardized gadget that you can stick in here and there and
that can work in a variety of different machines ... So these smaller
components are first to be identified because they are components
perhaps for lots of other tasks than the ones they’re used in.

Thus, there is room within a larger, contextually sensitive, address to his
materials (cf. the earlier-discussed contrast of Sacks’ starting point with that
of Searle) for the recognition and more formal description of particular
practices and sets of practices - here metaphorized as ‘gadgets’ or ‘machinery’
- which members can use in constituting coherent talk and specific lines of
action and interaction, and for an appreciation that some of these may operate
in a way substantially unqualified by the particulars of local context.

Recall again (cf the discussion above at pp. xx-xxii) the echoes in Sacks’
work in this period of some of the themes of work in generative grammar
(more accurately, an analytic model whose most lively embodiment at the
time was generative grammar, but which is surely not limited to that domain
of work). “ 4 The lectures for Fall 1965 were for a course whose catalogue title

23 It does not drop out as a theme of the lectures, however; cf the discussion at
pp. xxxvii-xxxviii below, and n. 26.

2-4 It is worth making explicit here that Sacks kept himself informed of

XXXVI

Introduction

was ‘Culture and personality.’ Whether or not he would otherwise have been
inclined to do so, it was perhaps this title which prompted some discussion by
Sacks of the notion ‘culture.’ In setting out the orientation of his examination
of the story told by a child, “The baby cried, the mommy picked it up,” Sacks
subsumed it, and the ‘machinery’ by which it was produced and heard, under
the notion ‘culture,’ of which he remarked, “A culture is an apparatus for
generating recognizable actions; if the same procedures are used for generating
as for detecting, that is perhaps as simple a solution to the problem of
recognizability as is formulatable’ (Fall 1965, Appendix A, p. 226, emphasis
in original.) 25 His description of what ‘the apparatus’ should do is strikingly
reminiscent of lines from early Chomsky, and seems directly targetted at
transformational grammar, but here, surprisingly, not at its principles - but
at its product: ‘We are going to aim at building an apparatus which involves
building constraints on what an adequate grammar will do, such that what an
adequate grammar will do, some of the things it will do, we are going to rule
out, and provide for the non-occurrence of’ (Fall 1965, Appendix A,
p. 229). Sacks’ undertaking here seems in important ways to be shaped by
the transformational grammar enterprise, albeit in a corrective stance toward
it. The stance seems to be something like the following. Given an undertaking
like the one generative grammar studies had seemed to set in motion, and
operating with similar sorts of goals (e.g., to generate all and only the
grammatical/acceptable sentences of a language), getting right results re-
quires looking at something other than just the linguistic or, even more
narrowly, the grammatical aspects of sentences or utterances. Not language,
but culture, is the key object and resource. And while such an enterprise was
understood by some ‘as ethnomethodology,’ by others it was seen as an
anthropological/cultural version of cognitive science (albeit along different

contemporary developments in a wide range of potentially relevant disciplines, and
was aware of what seemed to be ‘hot’ topics and ways of working. His work
recurrently speaks to such developments, sometimes explicitly, sometimes tacitly. He
is aware of, and responsive to, his intellectual ambience. The present account often
underscores such points of convergence and contrast - both with respect to the
ambience at the time Sacks’ work was being done and with respect to developments
at the time the present publication is being prepared. What may be of enduring
interest is the larger picture of the intellectual stances and developments at issue,
rather than the more transient excitements that pass over areas in ferment, even if
these substantially engage a generation of workers in a field.

25 On some readings, it is telling to compare this stance with Garfinkel’s account
of ethnomethodology (1967: 1), about whose studies he writes,

Their central recommendation is that the activities whereby members produce
and manage settings of organized everyday affairs are identical with members’
procedures for making those settings ‘account-able’ . . . When I speak of
accountable my interests are directed to such matters as the following. I mean
observable-and-reportable, i.e., available to members as situated practices of
looking-and-telling .

Introduction xxxvii

lines than those previously suggested by studies in ethnoscience and compo-
nential analysis).

There are various problem-types addressed and observations developed in
these lectures which seem to have a (sociological?) bearing on what came to
be called ‘cognitive science.’ Here I can mention only one of each.

First, observations. Both in Fall 1965 (lecture 7) and in Spring 1966
(lecture 18) Sacks comments on the differential ‘owning’ or control of certain
categories by different social groups, and the not uncommon asymmetry
between those to whom a category is applied and those who apply it. One
particular focus for this line of analysis is the pair of terms ‘adolescent’ and
‘hotrodder’ as applied to teenaged boys. ‘Adolescent’ is ‘owned’ by the
conventional adult society, and is deployed by its members (together with all
the commonsense knowledge or ‘conventional wisdom’ for which it is the
organizational locus in the culture) more or less without regard to the views
of those whom it is used to characterize. ‘Hotrodder’ (or, more recently,
‘punker,’ etc.), on the other hand, are categories deployed by their incum-
bents, and in ways often inaccessible to those who are not themselves
members. It is this relative independence from the ‘official’ or conventional
culture that led Sacks to term such categories ‘revolutionary’ (Spring 1966
lecture 18, and Sacks, 1979). There seems to be here a whole area of inquiry
which might be termed a sociology of cognition or a cognitive sociology quite
distinct from other usages of this term (cf. especially Cicourel, 1974). Insofar
as it involves the differential relevance of different category sets for the
cognitive operations of persons dealing with categories of persons, its
relevance to cognitive science seems transparent.

Second, problem-types. There is a form of problem which Sacks takes up
a number of times in the early lectures, each time on a distinct target, which
can be best characterized as an ‘analysis of the ordering of cognitive
operations’ (or the ordering of interpretive procedures). Two especially
brilliant instances of solutions to this problem-type occur in the Spring 1966
lectures. In lecture 1 1 (pp. 350-1) and again in lecture 21 (pp. 417-20), in
dealing with an instance of ‘intentional misaddress,’ Sacks wonders how the
co-participants in an interactional episode could have found who was being
addressed, since the address term employed by the speaker (“mommy”) did
not ‘actually’ apply to anyone present. He argues that, if they were finding
‘who is being addressed’ by finding to whom the address term referred, then
they would find no solution. Rather, he argues, they first use sequencing rules
to find whom the current speaker would properly be addressing, and they use
the product of that analysis in deciding how the address term is properly to
be interpreted. He is thus able to sort out the order in which these analyses are
conducted - first addressee, then address term - and it turns out to be just the
opposite from what one might have thought.

Another instance of the same problem is addressed in lecture 16 for Spring
1966. Here the object of interest is what is conventionally known as ‘the
possessive pronoun.’ Rather than taking a word like ‘my’ as indicating a
relationship of ‘possessing’ toward whatever it is affiliated to (which yields

Introduction

xxxviii

results in usages such as ‘my brother’ or ‘my teacher’ which are either
obviously faulty or in need of subsequent, and questionable, interpretation),
Sacks argues that a hearer/receiver must first determine that what ‘my’ is
attached to is a ‘possessable’ - the sort of thing which in that culture can be
possessed (rather than a category from a membership categorization device,
for example), in order to decide that ‘my’ is being used to claim possession.
Once again, an ordering of analyses - of cognitive operations - seems clearly
involved.

In both of these cases, the upshot of Sacks’ analysis is to reject as inadequate
the view that linguistic items determine the meaning or the force of an action,
and to insist instead that the cultural, sequential or interactional status of the
objects employed in the utterance shape the interpretation of the linguistic
item. 26

But for Sacks there was no in-principle ordering of what sorts of things one
consults first (e.g., the syntactic, semantic, sequential, interactional, etc.) and
no necessary priority, therefore, among the disciplines which study them.
Perhaps the first appearance of this problem-type is in lecture 4 for Fall 1965.
Here Sacks is discussing various forms of ‘tying rules,’ forms of talk (such as
indexical or anaphoric reference) which require a hearer to make reference to
another utterance to understand a current utterance, and which thus ‘tie’ the
utterances to one another. Encountering such usages of ‘that’ as “I decided
that years ago” or “That’s the challenge,” Sacks remarks that they present a
complication relative to other instances of tying procedures which he had
previously discussed, for such usages must be distinguished from the use of
‘that’ in, for example, “I still say though that if you take ...” Before
analyzing a ‘that’ for the sequential tying connection it makes to some other
(ordinarily prior) utterance, a hearer has to do a syntactic analysis to determine
that the ‘that’ is the sort which can tie back to some earlier component of the
talk. Here, once again, the sheer occurrence of an item (whether address term,
‘my,’ or ‘that’) does not determine what is to be made of it. But whereas in
the analyses previously discussed a linguistic analysis is contingent on prior
sequential, interactional or cultural analyses, here the sequential ‘tying’
analysis is contingent on a prior syntactic one. 27

26 Related discussions can be found throughout the lectures. For example, in the
Spring 1966 lectures: lecture 11, pp. 350-1 (re how sequential and interactional
organization controls the semantic and truth-conditional interpretation of an utter-
ance, rather than the opposite, which is the ordinary understanding); lecture 16,
p. 383; lecture 21, pp. 417-20; lecture 27, p. 451 (where sequential context is
shown to control the very hearing of a word); and lecture 29, pp. 461-3. See also the
earlier discussion at n. 23. Fuller discussion of the theme and the particular analyses
on which it rests must await another occasion.

27 Still, there is little doubt that the main thrust of analyses along these lines is that
the understanding of talk is, in the first instance, controlled by the hearer’s grasp of
the sequence in progress (or the sequential context more generally), rather than being
derived from the linguistic tokens. Cf., for example, the discussion in Spring 1966,
lecture 27, p. 451, where Sacks discusses the difficulty experienced by one participant

Introduction

XXXIX

Whatever the particulars, both these observations about control of
categorization structures and deployments and the problem-type addressed to
the ordering of cognitive or psycholinguistic or interpretive operations are
theoretically central to the responsibilities of a sociological, or more gener-
ally interactional, sector of what are now called the cognitive sciences. And
to the degree that the results of these inquiries inform and constrain our
understanding of how linguistic and category terms work, indeed can work,
their import goes well beyond the interactional domain which is their initial
locus.

The quasi-generativist themes in the Fall 1965 lectures, and in the 1964-5
lectures as well, co-exist with analyses of particular action types (‘how to do
action X’) based on empirical materials of talk, and co-exist as well with
analyses of sequencing and tying practices - also developed on empirical
materials, and addressed to the doing of conversation as an undertaking in its
own right. This variety of topics and approaches (and I have not mentioned
all the separate strands here) are, then, not a matter of stages in Sacks’
intellectual development over time. There are in these early lectures different
sorts of undertaking underway, differentially developed by Sacks, differen-
tially appealing to various segments of his professional readership, and
perhaps differentially susceptable to development by others, and, therefore,
differentially institutionalizable as a discipline. Surely, however, the drift of his
own subsequent work favored some of these initiatives over others.

VIII

If the lectures of Fall 1965 tilt in the direction of culture (whether incidentally
because of the course title or because it was central to Sacks’ preoccupations
at the time), the Spring 1966 lectures feature culture quite centrally. This was
the most extensively taped and transcribed of the lecture sets, and it is as rich
as anything in the materials assembled in these volumes. In its range - from
the empirical detail of the interactional materials to discussions of some of the
classic texts of social science and western culture - it gives the reader some
sense of the power of the mind at work here, of the nuanced sensitivity to
detail and of the scope of learning being brought to bear, and the distinctive
stance being developed through the conjunction of these resources. Here I can
touch only briefly on a few of the central themes of these lectures.

One theme, clearly part of the ‘culturalist’ motif of these lectures, and
surely not unrelated to the abiding preoccupation with ‘reflexivity’ and the
‘incarnate character of accounts’ central to the continuing development of
ethnomethodology in Garfinkel’s oeuvre, concerns the relationship between

in hearing something addressed to him which is acoustically accessible to everyone
else. He remarks that the party in question hears that turn by reference to the
sequence in which it occurs ‘so as to hear, indeed, a puzzle, when he could hear
something perfectly clear.’

xl

Introduction

‘commonsense knowledge’ and real world conduct or praxis on the one hand,
and between commonsense knowledge and ‘professional’ inquiry on the other
hand. This theme provides an opportunity as well to touch on the elements
of continuity and discontinuity in the orientation of Sacks’ work going back
to ‘Sociological description’ (1963).

Although there is no direct connection between the positions explored in
‘Sociological description’ and these lectures, there are echoes here, formal
similarities to aspects of the earlier paper. By ‘no direct connection,’ I intend
two observations. First, there is a substantial difference between what Sacks is
doing in the lectures and the hypothesized program of studies which Sacks
entertained in ‘Sociological description’ as a contrast with his depiction of
how contemporary sociological inquiries are conceived and carried through.
Second, there was no direct step-by-step theoretical development that led
from the position taken up in the 1963 paper to the directions pursued in
the lectures of 1964-6. On the other hand, I can only roughly suggest one
sort of observation I have in mind in suggesting ‘echoes’ and ‘formal
similarities.’

The central metaphor of ‘Sociological description’ was the so-called
‘commentator machine,’ a ‘device’ describable (from one point of view) as
composed of two parts - one which engages in some physical activity and
another which produces a form of language, understandable as a description
of what the first part is doing. Sacks entertains a variety of possible
formulations of this device, and the ‘proper’ understanding of the relationship
of its parts. The ‘doing part’ can be understood as a resource for coming to
understand what the ‘speaking part’ is saying. The ‘speaking part’ can be
understood as a description of what the ‘doing part’ is doing. The contraption
may be understood as two independent devices. And so on. For those views
in which the two parts do relate to one another, ‘discrepancies’ between the
parts can be variously understood: for example, as the ‘speaking part’ offering
inadequate descriptions of the ‘doing part;’ alternatively, as the ‘doing part’
malfunctioning and badly enacting the program set forth by the ‘speaking
part.’

With such a theme in the background consider just a few elements of the
first two lectures of Spring 1966 and some elements from the lectures of the
intervening year, 1964-5.

One of the central tasks which Sacks sets himself in the lectures on “The
baby cried’’ is providing an account of how recognizable activities are done,
and done recognizably. And in particular how the activity of ‘describing’ is
done, and done recognizably. The key starting point here is that descriptions
are recognizable, are recognizable descriptions, and are recognizable descrip-
tions without juxtaposition to their putative objects. Much of Sacks’ effort in
the early years of this analytic enterprise was given over to building an
apparatus that provided recognizable descriptions without reference (by real
life co-participants or by professional investigators) to what was putatively
being described. The ‘membership categorization devices’ introduced in
lectures 1 and 2 of Spring 1966, and the MIR device introduced in lecture

Introduction

xli

6 of the 1964-5 lectures (p. 4l) 28 are key elements in such an apparatus.
And the commonsense knowledge of the social world which is organized in
terms of these categories, ‘protected’ as it is ‘against induction’ (as Sacks
used to remark), provides for just such autonomously recognizable possible
descriptions. When some potential discrepancy is suggested between what is
provided for by the ‘knowledge’ organized around some category in a
categorization device and what is observably the case about some putative
incumbent of such a category, what may well be found (Sacks pointed out,
and this is part of what he meant by ‘protected against induction’) is not
the inadequacy of that ‘knowledge’ but rather the inadequacy of that
person as a member of the category involved, an inadequacy which that
person may feel and may seek to remedy.

Although vastly transformed (from a ‘doing part’ and ‘speaking part’ to
‘observable conduct’ and ‘recognizable description’), the problematics con-
cerning (1) the proper juxtaposition of the practical activities of social conduct,
(2) the commonsense knowledge of the mundane world and descriptive
practices resident in that world, and (3) the proper formulation of investiga-
tors’ stances and goals with respect to that world, persist from ‘Sociological
description’ through these lectures.

One component of these problematics is specially important throughout
these lectures, surfacing at the end in Spring 1966, lecture 33 but also central
at the beginning, and that is the relationship between commonsense knowl-
edge which investigators may share with those whose conduct is the object of
inquiry and the proper formulation of research questions, observations and
findings. Sacks begins the discussion of “The baby cried’’ with a number of
observations which he makes about the components of this little story, and
offers the claim that his audience would have made (perhaps did in fact make)
the same observations. But these are not sociological findings, he insists. They
are simply the explication of commonsense or vernacular knowledge. Rather
than constituting analysis, they serve to pose a research problem, namely, the
construction of an apparatus that would generate (or that has generated) such
observations, that would (in that sense) have produced them. And such an
apparatus would constitute findings.

Both parts of this analytic operation are important: making explicit the
understandings which common sense provides of the world which members
of the society encounter, including the conduct of others; and the provision of
something that can account for those understandings. And it is important to
keep them distinct and to insist on both.

Consider, for example, the notion of category-bound activities. It is in order
to address the observation that a report of ‘crying’ makes the category ‘baby’
(in the sense of a ‘stage-of-life’ category) relevant that Sacks introduces this
notion, and the proposal that the activity ‘crying’ is ‘bound’ to the

28 Sacks (ibid.) explains the term ‘MIR device’ by saying, “that is an acronym.
‘M’ stands for membership, T stands for inference-rich, and ‘R’ stands for
representative.”

xlii

Introduction

membership category ‘baby’ as one of the ‘stage-of-life’ categories in
particular. But the observation that “crying is bound to ‘baby’’’ is (like the
initial observations in the lecture) not a finding; it is merely the claimed
explication of a bit of commonsense knowledge. As such it is just a claim, and
cannot be simply asserted on the analyst’s authority. It has to be warranted
somehow, either by a test of it or by requiring it to yield some further pay-off
to analysis.

And this is what Sacks does with “crying being category-bound to baby.”
He immediately (lecture 1, p. 241) constructs a test of this category-
boundedness, even though (as he says) “it’s obvious enough to you, you
wouldn’t argue with the issue.” The pay-off, it will be recalled (lecture 1,
ibid), is not only the explication of ‘praising/denigrating’ as a test for the
category-boundedness of the action ‘crying,’ but an account for how to do
such recognizable actions as ‘praising’ or ‘deprecating’, research goals already
familiar from the 1964— 5 lectures and from elsewhere in the Spring 1966
opening lectures.

This stance is a basic and persistent one in these lectures. Elsewhere, for
example, Sacks insists on testing the claim that the categorization device
‘therapist/patient’ is ‘omni-relevant’ in the group therapy sessions which
supply the data for most of these lectures (Spring 1966, lecture 6, p. 315;
lecture 29, pp. 462-3; then again in Winter 1967, February 16, and Spring
1967, lecture 14), although this claim can be treated as no less ‘obvious.’ To
be sure, when he has recently made the point, Sacks sometimes asserts a
claimed category-bound activity without carrying through a test or deriving
a further finding (e.g., lecture 4, p. 302), but there can be little doubt that the
principle is basic - commonsense knowledge cannot properly be invoked as
itself providing an account, rather than providing the elements of something
to be accounted for. 29 In my view, Sacks abandoned the use of ‘category-
bound activities’ because of an incipient ‘promiscuous’ use of them, i.e., an
unelaborated invocation of some vernacularly based assertion (i.e., that some
activity was bound to some category) as an element of an account on the
investigator’s authority, without deriving from it any analytic pay-off other
than the claimed account for the data which motivated its introduction in the
first place.

The editorial effort to combine and blend largely overlapping treatments of
the same material, which has prompted the inclusion of lectures delivered
during the following term in Fall 1966, here in the Spring 1966 set (e.g.,
lecture 04. a), brings into relief certain shifts in analytic focus which
accompanied a return by Sacks to the same empirical materials. Only two of

29 See, for example, lecture 04. b, p. 295, here included with the Spring 1966
lectures, though actually delivered later, in Fall 1966: ‘. . . it is our business to
analyze how it is that something gets done, or how something is ‘a something,’ and
not to employ it.’

This theme - as represented, for example, in the phrase introduced by Garfinkel,
‘commonsense knowledge as topic and resource’ - is, of course, central to
ethnomethodology.

Introduction

xliii

these shifts can be taken up here, and only for a brief mention.

As remarked earlier, the analytic task set front and center in the initial
lectures for Spring 1966 was “how recognizable actions get done and get
done recognizably.” The first two lectures address those questions to the
actions ‘doing describing’ and ‘doing storytelling.’ (The third lecture, omitted
here because of its availability in a published version as ‘Everyone has to lie’
took up the issue of ‘doing a recognizably true statement.’) Lectures 04. a and
04.b, interpolated here from Fall 1966, have a different analytic focus -
observing and establishing orderliness - but lectures 4, 5 and 6 (delivered in
the Spring term) continue the ‘recognizable actions’ theme (doing recogniz-
able invitation and rejection) and reproducible methods for accomplishing
recognizable actions.

At the same time there is an apparent shift toward the invocation of a kind
of evidence that was to assume an increasingly central place in Sacks’
conception of how to ground an argument or an observation. In lecture 4
(from the Spring) he proposes that, in order to establish that “we were in an
automobile discussion” is doing a recognizable invitation, it is necessary not
only to agree that it “looks like an invitation” but to show “how that’s so”
(p. 301) with the description of a method for doing invitations that works for
the instance at hand. This echoes the stance of lectures 1 and 2.

In lecture 04. a (pp. 286-7, 288-9) from the Fall 1966 term, Sacks offers
as evidence that some earlier talk was attended by others than its overt
interlocutors, and as evidence that it constituted a recognizable introduction,
the prima facie evidence afforded by a subsequent speaker’s talk. Specifically,
he notes, that when Ken responds to the utterance of his name by the
therapist Dan not with “What” (as in an answer to a summons), indeed not
with an utterance to the therapist at all, but with a greeting to the newly
arrived Jim, he shows himself (to the others there assembled as well as to us,
analytic overhearers) to have attended and analyzed the earlier talk, to have
understood that an introduction sequence was being launched, and to be
prepared to participate by initiating a greeting exchange in the slot in which
it is he who is being introduced.

There is a shift here in analytic stance and procedure, from the analyst’s
understanding as initial point of departure on the one hand to the co-
participant’s understanding as initial point of departure on the other.

In the former mode, the analysis begins with an asserted convergence of
interpretations and recognitions by the analyst and the analyst’s audience (for
example, that something is a story, that ‘the mommy’ is ‘the mommy of the
baby,’ that an utterance is doing an invitation, and so on). It proceeds by the
provision of a methodical basis for both that convergence of understandings
and the convergence between the ‘understanders’ and the producers of the
to-be-understood ‘in the data.’ In the latter mode, analysis begins with an
asserted observation (that not-overtly-engaged participants are attending,
and, indeed, are obligated to attend to the talk), and then immediately
grounds that observation in subsequent conduct by the co-participants in the
episode being examined. That conduct is taken as displaying the product of

xliv

Introduction

their orientation to, and understanding of, the setting and what has been
transpiring in it. The site of analysis is located in the setting of the data at the
outset. And further: the analysts’ so treating the conduct of the participants
is itself grounded in the claim that the co-participants so treat it.

This contrast in stance and procedure is visible in this publication of the
lectures only briefly, by virtue of the juxtaposition of the material from Spring
and Fall 1966. What is seen only in lecture 0.4a-b here is seen increasingly
thereafter, starting with the Winter 1967 lectures in the present volumes. Of
course, this shift does not entail any abandonment of the commitment to
provide an account for how the recognizable outcome - whatever sort of
object it may be - is produced, although the form such an account might take
does change over time. The subsequently developed description of the
turn-taking organization, for example, is offered as a procedural account for
how a substantial collection of observable achievements of ordinary talk are
methodically produced by the co-participants.

What I have referred to as the ‘culturalist’ tenor of the Spring 1966
lectures is set in the first of its lectures, when Sacks sums up his initial gloss
of the understanding of “The baby cried . . .’’as indicative of “the operation
of the culture" as “something real and something finely powerful” (Lecture
1(R), pp. 245-6, emphasis supplied). The analysis of the membership
categorization device and of the commonsense knowledge organized by
reference to its categories is, in its fashion, an analysis of culture - “an analysis
of some culture,” as Sacks puts it (lecture 30, p. 469, emphasis supplied).
Throughout these 34 lectures (cf. especially lectures 13, 16-21, 24-25 and
3 1 and the appended manuscript ‘On some formal properties of children’s
games’) may be found treatments of various forms and artifacts of ‘culture’
in at least that anthropological sense in which it refers to the categories
through which ‘reality’ is grasped. Among these forms and artifacts are the
categories of persons making up a society and its world and who is entitled
authoritatively to ‘administer’ those categories (lecture 13), notions of
possession and possessables, the constitution of observations and descriptions,
measurements systems (lecture 24), games (lectures 13 and 31 and ‘On some
formal properties . . .’), conceptions of danger and their bearing on differen-
tially accomplishing such actions as warning and challenge (lecture 10, 12)
etc. A kind of socio-cultural semantics is involved, and a largely anthropo-
logical literature is invoked, reflecting Sacks’ engagement with then-
contemporary work in so-called ‘ethnoscience.’ 30

30 Cf. Sacks’ contrast of his own way of working on such matters with the
then-mainstream approaches to ethnoscience, for example, with regard to ‘measure-
ment systems,’ the discussion at lecture 24, p. 436, where the contrast is almost
certainly with the work of Berlin and Kay (1969, but circulated in mimeo earlier) on
color terms.

Although ethnoscience is in point for this particular reference, Sacks’ reading in,
and use of, the anthropological literature was very broad indeed - both in ‘areal’
terms and in ‘approaches.’ What he most appreciated was some combination of
dense and acutely observed ethnography, tempered by a sharp theoretical intelli-

Introduction

xlv

All of these lectures provide rich materials for analysis and discussion, but
in the present context, a brief consideration of Sacks’ treatment of games may
serve to recall some of the relevant intellectual context for this sort of cultural
analysis, as well as to permit a brief consideration of a direction for the study
of culture and acculturation, including language acquisition, which deserves
fuller exploration than it has been accorded.

The most immediately relevant context for writing about games within
American social science in the mid-1960’s traces back to the invention of
‘game theory’ in 1944 by von Neumann and Morgenstern as a branch of
mathematics with overtly ‘social’ applications (the title of their book was
Theory of Games and Economic Behavior ), with its subsequent elaboration by
economists and others concerned with strategic thinking, most visibly in the
late 1950’s and early 1960’s, in authors such as Kenneth Boulding (1963),
and Thomas Schelling (1961). The analytic force of the metaphor propelled
it into the arena of discourse and interaction as well, the language of
constitutive rules playing a central role in Searle’s development of speech act
theory, for example, and strategic considerations entering psychology and
sociology through varieties of ‘exchange theory’ (e.g., Thibaut and Kelley,
1959, or Blau, 1964). More proximately to Sacks’ thinking, both Goffman
and Garfinkel had explored the game model or metaphor in their own work

- Goffman in his essay ‘Fun in games’ (in Goffman, 1961) and later in
Strategic Interaction (1969, but written in 1966-7), and Garfinkel in the
so-called ‘trust’ paper (Garfinkel, 1963), a paper from which he subsequently
distanced himself, refusing to include it in the collection of his papers in 1967,
Studies in Ethnomethodology.

One problem with the assimilation of game theory into social science was
in establishing the limits of its usefulness as a model of social reality, a concern
surely central to both Goffman’s and Garfinkel’s treatment of it. One central
objection is that ‘games’ fail as a basic model of social order much as ‘contract’
failed as a basic model in Durkheim’s discussion of ‘utilitarian’ social theory,
an element of Durkheim (and Parsons’ (1937) treatment of Durkheim)
especially emphasized by Garfinkel. In both cases, the ‘model’ - whether
‘contract’ or ‘game’ — is itself ‘an institution,’ a normatively constrained
organization of understandings and conduct, with its own constitutive
infrastructure. ‘Contract’ could not undergird social order because, as a legal
institution, it was itself undergirded by the social order it was invoked to
explain. So also would ‘games’ fail as models of social interaction, for the

gence, and informed by broad learning. I recall especially his appreciation of Hocart
and Elizabeth Colson, of Fortune and Edmund Leach, of Evans-Pritchard and Max
Gluckman. But less reknowned ethnographers were no less appreciated. His fondness
for ethnography crossed disciplinary boundaries, and he collected original issues of
the volumes produced by the founding ‘Chicago school’ of sociological field workers

- Nels Anderson, Paul Cressey, Franklin Frazier, Clifford Shaw, Frederic Thrasher,
Harvey Zorbaugh - and later sociological ethnographies such as Dollard (1937),
Drake and Cayton (1945), and, in a different vein, studies like Cressey (1953),
discussed earlier.

xlvi

Introduction

conduct of games and their constitution presumed an infrastructure of
interactional conduct, and an epistemic/ontological definition as a discrete
order of ‘reality,’ within which games constituted a separate domain of
activities. Such misgivings would surely have informed Sacks’ approach to
games from the outset.

It is noteworthy, in this regard, that Sacks focussed on childrens’ games, 31
and that one of his central preoccupations was to get at that very infrastruc-
ture by reference to which games, as a special class of events, also are
undergirded. Thus, both in lecture 13 (on the game ‘Button-button who’s got
the button’) and in the draft manuscript on children’s games appended to the
Spring 1966 lectures, games are treated not as models of or about social life
for the social scientist, but as training grounds for formal aspects of social life
in social life, i.e., as arenas within social life for kids’ learning of central
features of (the) culture, features such as the operation of membership
categorization devices, the management of appearance and emotional display,
etc. His treatment of children’s games aims to provide analytic particulars for
his claim (‘On some formal properties . . . , Spring 1966, Appendix A,
p. 502) that “Play then becomes an environment for learning and demon-
strating criterial matters in real world action.” Games provide models of social
life in social life for its initiates, and in that capacity can be looked to for
methodically central components of culture. In that regard, for example, such
a game-relevant contrast as ‘counting’ versus ‘not counting’ can provide
materials on which can be built such ‘ real-world ’ contrasts as ‘legal versus
illegal.’

Considerations of enculturation and ‘language acquisition’ provide an
especially provocative focus for a matter which Sacks raises, in the first
instance, rather more as a methodological point. Taking up the methodolog-
ical relevance of sampling, Sacks points out that it depends on the sort of
order one takes it that the social world exhibits. An alternative to the
possibility that order manifests itself at an aggregate level and is statistical in
character is what he terms the ‘order at all points’ view (lecture 33, p. 484).
This view, rather like the ‘holographic’ model of information distribution,
understands order not to be present only at aggregate levels and therefore
subject to an overall differential distribution, but to be present in detail on a
case by case, environment by environment basis. A culture is not then to be
found only by aggregating all of its venues; it is substantially present in each
of its venues.

Leaving aside the consequences for the methodology of professional
inquiry, consider the implication that . . . any Member encountering from
his infancy a very small portion of it, and a random portion in a way (the
parents he happens to have, the experiences he happens to have, the
vocabulary that happens to be thrown at him in whatever sentences he

3 1 Recall that this antedates by several years organized attention to play and games
in the social science community, as represented, for example, in the wide ranging
collection edited by Bruner, Jolly and Sylva (1976).

Introduction

xlvii

happens to get) comes out in many ways pretty much like everybody else, and
able to deal with pretty much anyone else’ (ibid., p. 485).

In such a view, one might conjecture, we have one, and perhaps the major,
theoretically available alternative to Chomsky’s argument that, given the
highly limited and ‘degenerate’ sample of a language to which first language
learners are exposed, most of language - the crucial part - must certainly be
innate; they surely could not be induced from the available ‘inputs.’

The alternative is to consider a culture — and language as one component
of culture - to be organized on the basis of ‘order at all points. ’ If culture were
built that way, then socialization and language acquisition might well be
designed accordingly, and require induction from just the ‘limited’ environ-
ments to which the ‘inductee’ is exposed. As Sacks writes (ibid., p. 485), “. . .
given that for a Member encountering a very limited environment, he has to
be able to do that {i.e., grasp the order] . . . things are so arranged as to
permit him to.” ‘Things’ here presumably includes the organization of
culture, the organization of language, the organization of learning, and the
organization of interaction through which the learning is largely done. What
such a view projects is the need for an account of culture and interaction - and
the acquisition of culture and language in interaction - which would
complement a ‘cognitive’ language acquisition device and innate grammar
much reduced from contemporary understanding. Studies relevant to such a
view have been pursued for the last two decades or so, but not necessarily
under the auspices of the theoretical stance toward culture which Sacks
projects here. The evidence for an order at all points’ view has accrued
throughout Sacks’ subsequent work and the work of others working in this
area.

IX

The sessions from Winter 1967 appear in various respects transitional. There
are returns to, and revisions of, themes initially discussed in earlier sets,
including 1964-5 lectures, and initial explorations of topics taken up in much
greater detail in subsequent terms. The discussion here can only touch on a
few of these themes.

It is in the session of March 2, 1967 that we find the first substantial
consideration of turn-taking in multi-party settings. Here, as elsewhere in the
lectures, a set of materials is treated lightly near the end of one term, and then
is taken up in much greater detail in the next. The single session devoted to
tum-taking in Winter 1967 is followed by seven lectures in Spring 1967 (the
lectures on turn- taking from that term are not printed here), and an extensive
run in Fall 1967. 32

32 Another ‘take,’ embodying a different stance toward the work, is presented in
the Fall 1968 lectures.

xlviii

Introduction

A good deal of this treatment seems to have been prompted by reflections
on the difference between the two-party talk discussed in the 1964-5 and Fall
1965 lectures on the one hand (for which the materials were drawn from
telephone calls to the Suicide Prevention Center), and, on the other, the group
therapy sessions (GTS) on which many subsequent lecture sets are based.

But the relevance of working with multi-person talk was not limited to the
issue of turn-taking alone. To cite but one other product of the juxtaposition,
the discussion at pp. 529-33 of the March 2 session is concerned with
‘derivative actions,’ i.e., what a speaker may be doing to a third party by
virtue of addressing a recipient in a certain way. This seemed to Sacks but one
indication of the need to take up multi-person materials apart from two-party
ones (p. 533). 33

There is a theme taken up in the February 16 session, and touched on again
on March 9 (pp- 543-6), whose relevance to contemporary concerns (both
then and now) may be worth brief development here. One way of
characterizing those concerns is the generic relevance of context to talk in
interaction.

The general question taken up is whether there is some way of formulating
or invoking the sheer fact of the ‘settinged’-ness of some activity, without
formulating or specifying the setting. The ‘solution’ which Sacks points to is
the use of indicator terms (e.g., ‘here and now’ or stable uses of ‘this’) to do
this, a usage which affords us evidence that it can, in fact, be done. Indicator
terms can be seen as a machinery for invoking an unformulated setting, for
referring to (categorially-) unidentifed persons, or taking note of unformu-
lated activities.

But where does this ‘question’ come from? Why is its solution of any
interest? The beginning of the discussion, of course, is given not by a question,
but by some observations which end up as the ‘solution.’ This was a common,
and recommended, analytic procedure for Sacks: begin with some observa-
tions, then find the problem for which those observations could serve as
(elements of) the solution.

The observations in point here concerned the use of such ‘indicator terms,’
terms whose special relevance for ethnomethodology had (under the name
‘indexical expressions’) already been developed and underscored by Garfinkel
(1967, passim). And the central observations here had come up in a train of
considerations with a quite different focus, along the following lines.

The discussion begins with the problems of the ‘professional’ analyst (i.e.,
the ‘conversation analyst,’ not the ‘therapist’ in the data) establishing the
categorization device ‘patient/ therapist’ as omni-relevant for the participants
(which would cast it as always-invocable - ‘on tap,’ so to speak - both by
participants and by analyst). One way of doing that analytic task is to
establish a formulation of the setting as ‘group therapy session’ as omni-
relevant. Sacks then observes that this is but one form of ‘formulating as a

33 Subsequently it turned out that derivative actions can be found in two-party
conversation as well (Schegloff, 1984 [1976]).

Introduction

xlix

such-and-such,’ and that this is something that Members do. When they do it,
it is consequential, that is, they are doing some possible action in doing the
formulation. (Recall the discussion in Spring, 1966 of “we were in an
automobile discussion’’ as a formulation of the topic as a such-and-such
which is consequential - which does a possible invitation.)

The question then is: is there some way of referring to the context, or
components of the context, without formulating the context (or persons or
actions in it) as such-and-such - without, therefore, potentially doing the
actions which such a formulation might do. (Note that this can be a
consideration both for members/participants-in-the-interaction and for pro-
fessional analysts: for members so as to avoid doing the potential actions and
the responses they would engender in the interactional setting; for analysts
because it is precisely the escape from control by that interactional consequen-
tiality, from what otherwise constrains or ‘disciplines’ formulations, that
makes professional use of the lay device problematic).

It is in this context that the observation about the indicator terms finds its
resonance: terms like ‘here and now’ can invoke any present context and any
conception of scope-of-context (‘in this room,’ ‘in 20th-century America,’
etc.) without formulating it. And by requiring a recipient to provide its sense,
they recruit the recipient into the speaker’s project; they make the recipient
complicit in forming up its sense.

Several further brief comments will have to suffice here:

1 The observation that formulating does more than simply naming what
is formulated is focussed especially on ‘formulating what someone is doing’ in
the March 9 lecture (pp. 544-6), and it sounds a theme central to the paper
‘Formal properties of practical action’ (Garfinkel and Sacks, 1972), though
that paper was (according to Sacks) Garfinkel’s work. In this discussion in
Winter 1967 we see something of Sacks’ ‘take’ on similar issues, possibly one
source of discussions of this theme between them.

2 The considerations raised here (and when this theme is addressed
elsewhere) impose a constraint on discussions of ‘context’ and its bearing on
talk and action which has not been fully absorbed in the literature. The same
problems raised about the categorization of persons/members pertain:

the set of available characterizations is indefinitely extendable;
the selection of some one or more is potentially a way of doing something,
(i.e., is open to such understanding by others);
in actual interaction, such possible interpretation by interlocutors and the
responses they may offer in turn, can serve as a constraint on actually
selecting such a formulation;

the absence of such a constraint in the activities of professional analysts leaves
the grounds of such choices undisciplined, and therefore problematic.

The positivist solution to this problem (i.e., constraining the choice of
formulation by explanatory adequacy as attested by ‘evidence,’ leaves the

1

Introduction

actual orientations of participants out of the picture. Where ‘context’ is made
a central notion, these concerns will have continuing relevance.

3 The effort to cast ‘therapist/patient’ as an omni-relevant categorization
device has as one continuing relevance the concern with the bearing of gender
(and perhaps other of what Hughes (1971) used to call ‘master statuses’) on
talk-in-interaction. Were those who bring considerations of gender to bear on
all phenomena of interaction to take seriously the considerations touched on
just above, they might undertake to show that the pervasive relevance of
gender can be grounded in the demonstrably equally pervasive orientations to
it by participants to interaction. In effect, this would amount to showing that
the categories and terms of gender identification are omni-relevant for
interaction.

Sacks’ exploration of this issue in Winter 1967 is left unresolved. By the
time he comes to cast the indicator terms as ways of invoking the
settinged-ness of the interaction without formulating it, the problem of
establishing omni-relevance of either member-formulation or context-
formulation has been abandoned. In its place is the possibility of non-
formulation, of a kind of specific abstractness in treating the contexted
character of activity. But the exploration of omni-relevance is taken up again
in lecture 14 for Spring 1967 (cf. discussion below at pp. liii-livff.).

One theme from the 1964-5 lectures which reappears in the Winter 1967
lectures, reapplied to a related topic, is that of ‘direct’ versus ‘in-various-
ways-non-direct’ speaking; the topic to which it is now applied is ‘euphe-
mism’ (or what may be, once the data are examined, better termed ‘irony’).
Although the 1964-5 lectures asked over and over again ‘how to do X
without doing it overtly,’ the message here is that to ask why a euphemism
or ironic trope was used instead of a direct or ‘literal’ saying is to get the
question wrong. What Sacks is urging here (March 9, 1967, pp. 545-6) is
that the first-order consideration is not directness/indirectness or literalness/
figurativeness. Rather it is (for the speaker) a saying which displays its
relevance at that point in the talk, and (for the hearers) a saying such that
their understanding (their capacity to understand) ‘proves’ the utterance’s
relevance. The ‘norm’ is not, in the first instance, direct or literal reference,
but rather ways of talking that are locally adapted and can show local
relevance.

The first-order considerations are thus tying rules and other local connec-
tions between elements of the talk, rather than ‘saying it directly.’ The issue
of ‘directness/indirectness’ comes to the fore only with academic analysts
determined to understand the talk ‘in general,’ stripped of its local context.
For them what comes to identify a bit of talk, to constitute its re-referable
core, is its semantico-lexical content and perhaps its pragmatic upshot. With
that as the core, then various ways of realizing that central identity can come
to be formulated as more-or-less straightforward, direct, literal, or ‘tropic’ in
some respect. What was in situ a production tailored to the details of local
context is reinterpreted as a design for indirectness when local context is

Introduction li

stripped away and no longer accessible as the source of the utterance’s
design.

How is this line to be reconciled with the analysis in the 1964-5 lectures,
where just this question is asked - e.g., why seek out the other’s name without
asking for it directly? Perhaps this is one locus of development and change in
Sacks’ thought during this period. But it is also possible that when the
embodiment-of-indirection cannot be understood (by recipient, or by profes-
sional analyst) as an adaptation to the local context, then the question of why
the indirect rather than the direct may in fact be warranted and useful, and
in just those terms. 34

X

There are three predominants ‘casts’ to the lectures of Spring 1967.

As noted earlier, the first seven lectures (not published in this edition)
constituted the first sustained set on turn-taking, expanding the treatment in
the lecture of March 2 in the Winter 1967 set. This is a ‘sequential
organization’ cast.

Lectures 8-9, earlier treated in lecture 3 for Spring 1966 and subsequently
published as ‘Everyone has to lie’ (1975), have what might be termed more
of a ‘socio-logic’ cast - juxtaposing to what might appear ‘logical’ ways of
analyzing the conversational materials properly socio-logical ones.

From lecture 1 1 on, the materials take on the same flavor of
anthropological/cultural analysis that so heavily informs the Spring 1966 set.
This is largely the result of a focus on membership categories underlying talk
and relationships between those categories (their relative positionedness for
instance), notions of activities ‘bound to’ those categories, and the sorts of
commonsense ‘knowledge’ organized by reference to those categories (in the
manner of ‘Y do X,’ where Y is a category name, such as ‘women,’
‘freshmen,’ ‘politicians,’ etc.) Some of this material was organized into a draft
manuscript under the title, ‘On a device basic to social interaction,’ around the
time of writing of ‘An initial investigation . . .’ As introduced into these
lectures, much of the earlier statement seems to have been substantially
refined.

The discussion here will be limited to some reflections on the “Everyone
has to lie’’ analysis and on the reconsideration by Sacks of the matter of the

34 And Sacks does sometimes work on an utterance by addressing, what it prima
facie would be out of context, in a more-or-less ‘literal’ hearing, and with good
results; cf. Spring 1966, lecture 29 pp. 461—2, where he shows how various
components of the utterance “Usually there’s a broad in here” are neither produced
nor grasped in their ‘bare’ literal sense: e.g., ‘here’, means not ‘this place’ but ‘when
we are in {therapy] session;’ ‘a broad,’ means not ‘some woman’ but ‘the same
woman,’ indeed a particular same woman,’ and one who is a member of the group,
etc.

lii

Introduction

omni-relevance of formulations of setting and participants, earlier taken up in
Spring 19 66 and in Winter 1967.

At least one underlying source and rationale for the animating question
being addressed in the ‘exercise’ concerned with the assertion “everyone has to
lie” is formulated by Sacks (Spring 1967, lecture 8, p. 549) as “How could
we as social scientists go about saying about something that a Member said,
that it’s true.” It may be useful to ‘unpack’ the background for this question
at least partially. 35

As rhetoric as a core method and discipline for the analysis of what can be
said gradually became demoted in the intellectual hierarchy of western
culture, and logic developed an increasing hegemony, it brought with it an
increasingly exclusive preoccupation with ‘truth’ as the paramount feature of
assertions requiring definition and assessment. In part this concern was in the
service of ‘science,’ and its aims of establishing stable propositions about the
world whose truth could be established once and for all.

When attention began in the 20th century to turn to statements in
so-called ordinary language, the analytic apparatus available for use was that
of formal logic, and it was in part by virtue of the results of applying a formal
logic developed in the service of science and mathematics to ordinary language
that natural languages were found defective and the need for ‘formal
languages’ made compelling. But the goals of logic/science and ordinary
discourse are by no means the same, and the use of language in them may be
quite different. What is relevant to establishing the truth of a proposition in
science - and what might be ‘meant’ by ‘truth’ - may be quite different from
assessing the truth of a ‘commonsense assertion’ in ordinary circumstances. It
is this gap which, in part, Sacks is addressing.

Here, as elsewhere, Sacks’ exploration of this theme (the contrast between
‘common sense’ and ‘scientific’ procedures) is focussed on a class of terms
which is especially symbolic of logic — quantifiers. In ‘An initial investiga-
tion . . .’as well such a term had become a focus of analysis. There it was the
term ‘no one,’ in the claim by a suicidal person that they have “no one to turn
to,” and Sacks undertook to explicate how ‘no one’ is used, and used
‘correctly,’ given the ‘paradox’ that the assertion is made precisely in the
conversation in which its speaker has turned to ‘someone.’

‘Initial investigation . . .’ showed how “no one to turn to” was not belied
by having turned to someone for the conversation in which it was said because
‘no one’ had as its scope only certain categories of person; ‘no one’ was not
being used in some formal logical sense, as ‘no person.’ It was therefore
misguided to begin with a ‘logical’ understanding of the term, when that was
not the use being made of it in the production of the utterance.

In lectures 8 and 9, the quantifier under examination is ‘everyone.’ Again,
Sacks proposes not to begin with some sense of the term derived from logic
(some ‘strict usage’ as he puts it), and find how trivially to disprove the
assertion by showing that there is at least one person who does not have to lie.

35 A similar question is taken up in Spring 1966, lecture 26.

Introduction

liii

Rather, he proposes that we must investigate anew, and for its usage in
ordinary conversation, how a term like ‘everyone’ is constituted and used.

And more generally, assessing the truth of the assertion involves not just a
manipulation of truth conditions, but rather an explication of those practices
of talk-in-interaction which the assertion could reflect an orientation to, and
whose actual operation could be what is being invoked in the asserted claim.
In the context of this lecture, this refers to the contingencies of the ‘How are
you’ question, its privileges of occurrence, its types of relevant answer, and
how the further courses of action which its answers make contingently
relevant affect the choice of answers in the first instance (pp. 556ff.). By the
end of the discussion, this structure is generalized well beyond ‘How are you,’
and is used to specify where lying may be generically suspected, where
confessions of it will be readily believed, etc.

In any case, what emerges as criterial to the inquiry is not a logical analysis
of the component terms of the assertion and an assesment of their combina-
tion, but a social analysis of those contingencies of interaction which could
give rise to the condition which the assertion claims. The upshot here is to
blunt the prima facie application of ‘logical’ analysis as the first-order
consideration in much the same fashion as several of the Spring 1966 lectures
had the import of blunting the prima facie linguistic analysis of an utterance
(cf. above at pp. xxxvii-xxxix, the discussion of ‘the ordering of analyses’). In
both cases, the tools of linguistic and logical analysis are shown to have their
relevance and applicability constrained by, and contingent on, prior sequen-
tial, interactional and cultural specifications of the practices of talking
underlying production of the utterance.

What emerges is, then, a wholly different conception of what the analysis
of ordinary discourse should consist in. It is this result which is adumbrated
by asking at the outset how social scientists might go about assessing the truth
of what a Member says, and this which animates that question. 3

In lecture 14 (from p. 594 to the end of the lecture) Sacks again takes up
the question of the ‘omni-relevance’ of a category collection. In the discussion
of Winter 1967, the issue became redefined as invoking a context (and
potentially associated membership categories) without actually formulating
them - invoking the sheer fact of ‘settinged-ness’ (cf. above, pp. xlviii-1).
Here, the discussion remains focused on the possibility of omni-relevance.

What he means by ‘omni-relevance,’ Sacks says, is two-fold: ‘on the one
hand, there are some actions which, for their effectiveness [i.e. , to be
recognized as that type of action}, involve categorial membership in that
collection, and, on the other hand, until the course of action is ended, one
can’t rule out the further use of that collection.’ The elegant solution to the
problem of showing ‘therapist/patient’ to be omni-relevant in the empirical
materials under examination lies in noting that the effective doing of an
ending to the occasion requires reference to the status of one of the parties as

36 This sort of inquiry may be seen to inform the first paragraphs of lecture 1 1 as
well.

liv

Introduction

‘therapist.’ The point is made even more exquisite by ‘the therapist’ actually
only hinting at the ‘session’s’ closure, and one of the more experienced
patients interpreting that hint for a new patient.

What is key to the solution is its focus on the efficacy of the utterance in
implementing the action of initiating the ending of the session and the
non-contingency of that action. Other actions could be understood to activate
the relevance of the categories germane to their efficacy, but those categories
might not on that account alone necessarily be claimable as omni-relevant.
But accomplishing an ending is, first, a non-contingent occurrance for the
occasion (the issue is not whether it will be done, but when), and therefore
prospective, i.e., relevant even before an action might invoke it. It is this
non-contingent prospective relevance of an action — an action which itself
makes a membership category relevant - which grounds the argument for
omni-relevance here. 37

This lecture affords an especially clear example (as Sacks’ own lead-in
makes clear) of one form which his kind of theorizing took. It regularly began
with an observation about the particular materials being examined (an
observation, of course, commonly informed by his prior work and wide
reading). That observation might then be ‘developed:’ its terms being given
an ‘anterior’ development, i.e., he would find and explicate what his own
initiating observation could be seen, on reflection, to have presupposed; those
presuppositions might well be more ‘observations,’ and more consequential
ones. That package of observations might be followed up through discussion
of matters in the literature which they touched off, through exploring purely
formal kinds of logics they suggested, purely ‘theoretical’ possibilities they
seem to entail, etc. But, recurrently, these ‘theoretical’ developments would be
brought back to empirical materials - either what had initiated the whole line,
or other materials which the line of theorizing brought to mind. It was in this
sense that the effort was prosecuted to put theorizing at every point under the
control of empirical materials.

The actual presentations sometimes obscured this way of working. In
lecture 14, for example, Sacks begins with what appear to be very abstract
considerations about applying categories to partition a population, and the
relationship between the partitionings yielded by different category collec-
tions. This then is putatively ‘applied’ to the material at hand, in the analysis
of ‘teenager/adult’ as a ‘cover’ collection preserving partitioning constancy
with ‘patient/therapist;’ and in the covering of ‘patient/observer’ with
‘performer/audience.’ It was initially an observation about the latter - re the
utterance “Testing” (p. 593) in particular - which motivated much of this
line. Of course, the most extensive such reversal of order of discovery and order

37 How Sacks’ line of argument might bear on a claimed omni-relevance of gender
(to re-pose an issue earlier discussed) is unclear. At the least, the constraint of “until
the course of action is ended one can’t rule out the further use of that collection”
requires working out in any occasion being examined, specifically what ‘the course of
action’ can be taken to be.

Introduction

Iv

or presentation is the paper ‘An initial investigation . . . in which the
originating observation was about “no one to turn to,’’ the serious exploration
of which led to formulating it as the result of a search procedure, which
required formulating the terms of the search and the categories by reference
to which it is conducted, etc. It was with the last of these that the paper itself
began.

XI

The Fall 1967 lectures turned out to be the last at UCLA. Sacks’ teaching
during the Spring 1968 term was in seminar format, although he did offer
sustained presentations on occasion, and these are included in the present
volumes. And by Fall 1968 Sacks had moved to the University of California,
Irvine (although there is no reason to think the prospect was already known
at the time of the Fall 1967 lectures, or informed their delivery).

These lectures include the first extended treatment of turn-taking presented
in these volumes, although the first seven lectures for Spring 1967 (not
printed here) represented Sacks’ actual first effort on this scale. The Fall 1967
lecture set is the only one in which Sacks offered extended treatments of both
turn-taking organization and tying structures. Tying structures are discussed
in several earlier lecture sets, but not again after Fall 1967. And the
discussions of identification and categorization to which Sacks returned several
times in the lectures preceding Fall 1967 are not taken up here, and
henceforth reappear only sporadically and for much briefer treatment.
Sequential organization increasingly dominates the agenda of Sacks’ lectures,
including expanding discussions of turn-taking, of sequence structure and
adjacency pairs, of overall structural organization, of story-telling organiza-
tion, etc.

If the Spring 1966 lectures were especially ‘anthropological’ in orientation,
then the Fall 1967 lectures are especially oriented to linguistics.

This note is sounded early, when in the initial lecture, a general introduc-
tion, Sacks (pp. 622-3) projects the preoccupations of the course with
‘sequential analysis’ (though not under that name), which he introduces by
remarking that ‘ . . . the discoverable aspects of single utterances turn out to
be handleable - perhaps handleable only - by reference to sequencing
considerations . . . ,’ and declaring his interest in “. . . how it is that sequenc-
ing considerations turn out to be implicative of what happens in a given
utterance.”

“Linguistics,” by contrast (he argues), “is that study of the utterance which
involves detecting those features of it which are handleable without reference
to such considerations as sequencing; i.e., without reference to that it has
occurred in conversation” (ibid.).

One question, then, is whether “there is the possibility of. . .a fully
comprehensive, coherent linguistics without such matters.” Another is how
such study of single utterances can be “brought into alignment with what we

lvi

Introduction

know about sociology and anthropology. And if not, what then?” 38

Recurrently throughout these lectures Sacks brings the results of a line of
analysis or argument into juxtaposition with the main thrust of contemporary
linguistic theory and analysis (i.e. , of the early to mid-1960s). One result is
the sketching of whole orders of observable regularity and apparent normative
organization which have largely, in some cases entirely, escaped the notice of
the main thrust of the contemporary study of ‘language.’ In some respects,
this is undoubtedly related to the ambition of modern linguistics (tracable at
least to de Saussure) to transcend particular contexts and media of language
use - not only social and cultural settings, but also oral and written
embodiments - so as to describe an underlying, presumably invariant,
linguistic code. The attention to sequential organization - an order of
organization seemingly inescapable in the effort to understand and describe
actual, naturally occurring talk in interaction - forcefully belies the premise of
the currently dominant commitments of linguistics. Running through both
the Fall 1967 lectures and the presentations of Spring 1968 are several
recurrent themes, whose central upshot is:

How sequential considerations necessarily inform or bear on the construction
and understanding of single utterances;

How understanding of some talk is regularly displayed by its recipients; and
What that has required of recipients, and how those requirements are
formative of their talk in turn.

These themes are returned to persistently, almost compulsively, and they are
considerations of a ‘foundationalist’ sort - that is, they go to the matter of
what foundations a discipline of language must be understood to rest on.

Sacks has seemed to some to have abandoned his commitment to
contextually-sensitive analysis in turning to the study of sequential structure,
and turn-taking in particular. But in insisting on the decisive relevance of
sequential organization as furnishing the most proximate reference points of
context, Sacks showed the consequences of disattending the fact that language
was being used in a medium which was inexorably temporal and interac-
tional. The results of these explorations of sequential context offer, in their
own way, as sharp a contrast to formal linguistic analysis as did Sacks’ earlier
explorations in the 1964-5 lectures offer a contrast with Searle’s efforts at
context-free speech act theorizing (cf. above, pp. xxiv-xxix).

38 Later (for example, in a letter to me in 1974) Sacks seems to have taken a
different tack, namely, that a systematic discipline might not be buildable on the
analysis of single utterances, or single instances of other units or occurrences, but that
large amounts of material might be needed. At the time of his death, we had just
begun a large-scale investigation of ‘next turn repair initiators’ which was going to be
an exploration of that sort of undertaking. This subsequent development, of course,
in no way blunts the impact which Sacks produced by asking what was to be made
of the single utterance or the single sequence or the single exemplar of anything to be
analyzed, and the detailed findings which this way of working led him to.

Introduction

Ivii

Appreciation of the recurrent linguistic orientation in many of these
lectures should not be allowed to obscure the range and variety of matters
taken up in them, and the diversity of the intellectual resources being called
upon from many different traditions of inquiry. One case in point must
suffice.

In lecture 6 for Fall 1967 Sacks returns to a point which had come up in
earlier sets of lectures, concerning the inclusion in analysis of things which did
not happen, here offering as one special relevance of ‘next-speaker selection
techniques’ and ‘paired utterances’ (the later ‘adjacency pairs’) that they
provide enhanced analytic leverage for speaking of something being absent -
e.g., the utterance of an unresponsive selected next speaker, or the absence of
a responsive paired utterance. The problem of warranting claims about
‘absences’ has resonated to many corners of the conversation-analytic domain
of issues. Then Sacks adds (p. 670):

A way, perhaps, to develop a notion of ‘absence’ involves looking to
places where such a notion is used and attempting to see whether there
are various sorts of relevance structures that provide that something
should occur. Parenthetically, I’ll give as a rule for reading academic
literature, that whenever you see somebody proposing that something
didn’t happen - and you’ll regularly find, e.g., sociologists, anthropol-
ogists, or historians particularly, saying that something didn’t happen,
something hadn’t been developed yet - that they’re proposing that it’s
not just an observation, but an observation which has some basis of
relevance for it.

Sacks’ interest in the matter of ‘absences’ antedates his work with conversa-
tional materials. He had taken a special interest in an observation of Max
Weber’s that some aspect of ancient Middle Eastern history was to be
understood by reference to the fact that (as Sacks would put it in conversation)
“that was before the appearance of the horse as an instrument of warfare.” 39
The issue this posed was, how could something be the consequence of
something which had not happened yet? Clearly some set of relevancies to
which the theorist was oriented informed this way of thinking.

And, earlier yet, I recall a conversation at the Law and Society Center in
Berkeley in 1962-3 (involving Sacks, a Marxist graduate student in sociology
from Argentina and myself) in which the discussion lingered on ‘explanations’
for the absence of revolutions founded on the Marxist notion of ‘false
consciousness.’ At issue were both the theoretical status of observations

39 Weber (1952: 6, emphasis supplied):

Because the nature of military and administrative technology of the time
precluded it, before the seventeenth century BC, a lasting political conquest
was impossible for either of the great cultural centers. The horse, for instance,
while not completely absent, at least, not in Mesopotamia, had not as yet been
converted into an implement of special military technique.

lviii

Introduction

concerning the non-occurrence of revolution, and the reliance, in the concept
of ‘false consciousness,’ on a stipulated account by the theorist/analyst of
what the ‘real’ interests of the proletariat were, a correct appreciation of which
was ‘absent’ from their (i.e. , workers’) understanding of the world. What
made those ‘understandings’ relevant, such that not sharing them amounted
to their ‘absence,’ and rendered other beliefs of the working class to be ‘false
consciousness,’ with sufficient explanatory power to account for the absence of
revolution?

So when Sacks refers in lecture 6 to a ‘rule for reading academic literature,’
there is specific background informing the line he is recommending. Having
initially engaged this issue in the social science literature, Sacks came to find
it illuminated in his engagement with interactional materials. For the
underlying ‘logic’ was, although encountered in the first instance in academic
materials, but an aspect of ‘commonsense’ or ‘practical’ theorizing which had
been incorporated in professional social science theorizing.

Eventually Sacks pursued this matter with a variety of interactional
materials. For example, in one of the 1964-5 lectures he remarks on the
special intimacy and power of a line reportedly addressed to a beloved in
explanation of some past bit of biography, “That was before I met you, and
I was lonely then.’’ Here again a ‘state-of-the-world’ is explained by
something that had not yet happened, in a powerful display of retroactive
relevance.

So these lectures of Fall 1967, however oriented to exploring their interface
with contemporary linguistics, retain their grounding in social (even ‘socio-
logical’) and cultural analysis. Indeed, it is at the meeting point of these
disciplines that the analytic action of these lectures is situated.

XII

This volume presents roughly the first half of those lectures which Sacks chose
to tape record and have transcribed. The introduction to this point has
attempted to provide some thematic overview of these lectures, and some-
what more detailed background and exploration of a few selected issues.

This effort at an overview has been truly daunting, indeed, beyond my own
capacities at the present time. Part of this may surely be traced to my own
shortcomings. But, for the most part, it reflects rather the extraordinary
richness and multi-facetedness of Sacks’ corpus. In its variety, depth, and
freshness of vision it defies domestication into convenient guidelines to a
reader. At least part of this derives from the methodological character of
Sacks’ initiative - the new way of working he introduced. Starting out with
a commitment to lay bare the methodicity of ordinary activities, and with his
talent for seeing in singular occurrences the structural elements of which they
were formed and composed, a world of data which refreshed itself every
moment more than a legion of Sackses could ever make a dent in provided
a virtual infinity of opportunities for new observations, and new orders of
observation.

Introduction

lix

Not that it was easy! Sacks often complained about how hard the work
was, and that it did not seem to get easier. He spoke in the early 70s of giving
it up and working on something less demanding. The problem was, he
observed, the need to see “around the corner,” to penetrate through the
blinders of the implacable familiarity of the mundane materials with which
we worked, and the commonsense models and expectations derived from a
social science which had never addressed itself to the simple observational
tasks of a naturalistic discipline in which such models ought to have been
grounded in the first instance. If we were to try to build a discipline, we
needed to be able to be freshly open to what could be going on in any given
piece of interaction, and to how activities and conduct could possibly be
organized. And it was hard to say which was more difficult - to see clearly
what was going on in some bit of material, or to figure out how to build from
such observations and analyses a worthy discipline. And, of course, these were
not independent orders of task - for how to address the empirical materials
was always being informed by the direction in which it appeared a discipline
might be pursued, and one surely wanted the character of the discipline to be
shaped centrally by one’s sense of how social activities were actually
organized.

In any case, the main line of engagement for Sacks was in directly taking
up particular occurrences, particular bits of tape and transcript. And in
leaving as open as he could what there was to be noticed about that bit of
occurrence, what there was to be learned from it, what we might get to see the
importance of for the first time. And this insistence on freeing each next
engagement with data from the past - not only the past of the social sciences,
but also past work of this sort, including (especially) his own — while still
allowing it somehow to inform analysis is what allowed each new fragment of
data, each next look back at an old fragment of data, to provide a possible
occasion of discovery. Although the sorts of things which emerged (however
rich and multifaceted) were constrained by the particular metier of his mind,
their range was truly astounding. They overflow efforts to contain them and
package them for overview.

References

Anderson, Alan Ross and Moore, Omar Khayyam, I960. Autotelic folk models.
Sociological Quarterly, 1, 4, 203-16.

Atkinson, J. Maxwell and Heritage, John C. (eds), 1984. Structures of Social Action
(Cambridge: Cambridge University Press).

Austin, John L., 1962. How to Do Things With Words (Oxford: The Clarendon Press).
Barker, Roger G. and Wright, Herbert F., 1951. One Boy’s Day (New York: Harper and
Bros.).

Barker, Roger G. and Wright, Herbert F., 1954. Midwest and its Children (Evanston,
III.: Row Peterson and Co.).

Berlin, Brent and Kay, Paul, 1969. Basic Color Terms: their universality and evolution
(Berkeley and Los Angeles: University of California Press). Previously circulated as a

lx Introduction

Working Paper by the Laboratory on Language Behavior Research, University of
California, Berkeley).

Blau, Peter, 1964. Exchange and Power in Social Life (New York: John Wiley).

Blumer, Herbert, 1939. An Appraisal of Thomas and Znaniecki’s The Polish Peasant in
Europe and America (New York: Social Science Research Council).

Boulding, Kenneth, 1962. Conflict and Defense: a general theory (New York: Harper).

Brown, Penny and Levinson, Stephen C., 1978. Universals in language usage: politeness
phenomena. In Esther Goody (ed.), Questions and Politeness: strategies in social
interaction (Cambridge: Cambridge University Press), 56-310.

Bruner, Jerome S., Jolly, Alison, and Sylva, Kathy, (eds), 1976. Play: its role in
development and evolution (New York: Basic Books).

Button, Graham and Lee, John R. E. (eds), 1987. Talk and Social Organization
(Clevedon: Multilingual Matters).

Chomsky, Noam, 1957. Syntactic Structures (The Hague: Mouton).

Chomsky, Noam, 1965. Aspects of the Theory of Syntax (Cambridge: MIT Press).

Cicourel, Aaron, 1974. Cognitive Sociology : language and meaning in social interaction
(New York: Free Press/Macmillan).

Cressey, Donald, 1953. Other People's Money (Glencoe, III.: The Free Press).

Dollard, John, 1937. Caste and Class in a Southern Town (New Haven: Yale University
Press).

Drake, St. Clair and Cayton, Horace, 1945. Black Metropolis (New York: Harcourt
Brace & Co.).

Garfinkel, Harold, 1959. Aspects of the problem of commonsense knowledge of social
structure. Transactions of the Fourth World Congress of Sociology, 4, 51-65.

Garfinkel, Harold, I960. The rational properties of scientific and commonsense activities.
Behavioral Science, 5, 72-83.

Garfinkel, Harold, 1963. A conception of, and experiments with, ‘trust’ as a condition
of stable concerted actions. In O. J. Harvey (ed.). Motivation and Social Interaction
(New York: Ronald Press), 187-238.

Garfinkel, Harold, 1964. Studies of the routine grounds of everyday activities. Social
Problems, 11, 225-50.

Garfinkel, Harold, 1967. Studies in Ethnomethodology (Englewood Cliffs, NJ: Prentice
Hall).

Garfinkel, Harold and Sacks, Harvey, 1972. On formal structures of practical actions. In
John C. McKinney and Edward A. Tiryakian (eds), Theoretical Sociology (New York:
Appleton-Century-Crofts) .

Goffman, Erving, 1959. The Presentation of Self in Everyday Life (New York: Doubleday
Anchor).

Goffman, Erving, 1961. Encounters: two studies in the sociology of interaction (Indiana-
polis: Bobbs-Merrill).

Goffman, Erving, 1963. Behavior in Public Places (New York: The Free Press).

Goffman, Erving, 1964. The neglected situation. In John J. Gumperz and Dell Hymes
(eds). The Ethnography of Communication , special issue of the American Anthropologist,
66, 6, part 2, 133-6.

Goffman, Erving, 1969. Strategic Interaction (Philadelphia: University of Pennsylvania
Press).

Goffman, Erving, 1971. Relations in Public (New York: Basic Books).

Goffman, Erving, 1981. Forms of Talk (Philadelphia: University of Pennsylvania Press).

Halliday, Michael A. K., and Hasan, Ruqaia, 1976. Cohesion in Discourse (London:
Longman).

Havelock, Eric A., 1963. Preface to Plato (Cambridge, Mass.: Harvard University Press).

Hill, Richard J, and Crittenden, Kathleen Stones, (eds), 1968. Proceedings of the Purdue
Symposium on Ethnomethodology, Institute Monograph Series No. 1 (Institute for the

Introduction

lxi

Study of Social Change, Department of Sociology, Purdue University).

Hughes, Everett C., 1971. Dilemmas and contradictions of status. In The Sociological Eye:
selected papers (Chicago and New York: Aldine and Atherton), 141-50. Reprinted
from The American Journal of Sociology, 50 (1945).

Jefferson, Gail, 1983. On embedded and exposed correction in conversation. In Studium
Linguistik, 14, 56-68. Reprinted in Graham Button and John R. E. Lee (eds). Talk
and Social Organization (Clevedon: Multilingual Matters), 86-100.

Jefferson, Gail, 1989. Editor’s notes. Human Studies, 12, 395-404.

Kaufmann, Felix, 1944. The Methodology of the Social Sciences (New York: Humanities
Press).

Kneale, William and Kneale, Martha, 1962. The Development of Logic (Oxford: The
Clarendon Press).

Labov, William and Waletsky, Joshua, 1966. Narrative analysis: oral versions of
personal experience. In J. Helm (ed.), Essays on the Verbal and Visual Arts (Seattle:
University of Washington Press), 12-44.

Lakoff, Robin, 1973. The logic of politeness: or minding your p's and q’s. Proceedings of
the Ninth Regional Meeting of the Chicago Linguistic Society, 292-305.

Levinson, Stephen C., 1983. Pragmatics (Cambridge: Cambridge University Press).

Opie, Iona and Opie, Peter, 1959. The Lore and Language of Schoolchildren (Oxford: The
Clarendon Press).

Parry, Milman, 1971. The Making of Homeric Verse: the collected papers of Milman Parry,
edited by Adam Parry. (Oxford: The Clarendon Press).

Parsons, Talcott, 1937. The Structure of Social Action (Glencoe, Illinois: The Free Press).

Pomerantz, Anita, 1980. Telling my side: ‘limited access’ as a ‘fishing’ device. Sociological
Inquiry, 50, 3-4, 186-98.

Psathas, George (ed.), 1979. Everyday Language: studies in ethnomethodology (New
York: Irvington Press).

Sacks, Harvey, 1963. ‘Sociological description.’ Berkeley Journal of Sociology, 8, 1-16.

Sacks, Harvey, 1972a. An initial investigation of the usability of conversational data for
doing sociology. In David N. Sudnow (ed.), Studies in Social Interaction (New York:
The Free Press), 31-74.

Sacks, Harvey, 1972b. On the analyzability of stories by children. In John J. Gumperz
and Dell Hymes (eds), Directions in Sociolinguistics (New York: Holt, Rinehart and
Winston), 325-45.

Sacks, Harvey, 1972c. Notes on police assessment of moral character. In David N.
Sudnow (ed.), Studies in Social Interaction (New York: The Free Press), 280-93.

Sacks, Harvey, 1973. On some puns, with some intimations. In Roger W. Shuy (ed.),
Report of the 23rd Annual Roundtable Meeting on Linguistics and Language Studies
(Washington, DC: Georgetown University Press), 135-44.

Sacks, Harvey, 1974. An analysis of the course of a joke’s telling in conversation. In
Richard Bauman and Joel Sherzer (eds). Explorations in the Ethnography of Speaking
(Cambridge: Cambridge University Press), 337-53.

Sacks, Harvey, 1975. Everyone has to lie. In Ben Blount and Mary Sanches (eds).
Sociocultural Dimensions of Language Use (New York: Academic Press), 57-80.

Sacks, Harvey, 1978. Some technical considerations of a dirty joke. In Jim Schenkein
(ed.). Studies in the Organization of Conversational Interaction (New York: Academic
Press), 249-70.

Sacks, Harvey, 1985. The inference-making machine: notes on observability. In van
Dijk, 1985, 13-22.

Sacks, Harvey, 1986. Some considerations of a story told in ordinary conversation.
Poetics , 15, 127-38.

Sacks, Harvey, 1987 [1973]. On the preferences for agreement and contiguity in
sequences in conversation. In Button and Lee, 1987, 54-69.

lxii Introduction

Sacks, Harvey, 1987. “You want to find out if anybody really does care” In Button and
Lee, 1987, 219-25.

Sacks, Harvey, 1988/89. On members’ measurement systems. Research on Language
and Social Interaction, 22, 45-60.

Sacks, Harvey, ScheglofF, Emanuel A., and Jefferson, Gail, 1974. A simplest systematics
for the organization of turn-taking for conversation. Language 50, 696-735.

Schegloff, Emanuel A., 1963. Toward a reading of psychiatric theory. Berkeley Journal of
Sociology, 8, 61-91.

Schegloff, Emanuel A., 1967. The First Five Seconds: the order of conversational openings.
PhD Dissertation, Department of Sociology, University of California, Berkeley.

Schegloff, Emanuel A., 1984 [1976]. On Some questions and ambiguities in conversa-
tion. In Atkinson and Heritage, 1984, 28-52, [first appeared in Pragmatics Microfiche ,
1976, 2.2, D8-G1.]

Schegloff Emanuel A., 1988. Goffman and the analysis of conversation. In Paul Drew
and Tony Wooton (eds), Erving Goffman: perspectives on the interaction order (Cam-
bridge: Polity Press).

Schelling, Thomas, I960. The Strategy of Conflict (Cambridge, Mass.: Harvard Univer-
sity Press).

Schenkein, Jim (ed.), 1978. Studies in the Organization of Conversational Interaction (New
York: Academic Press).

Searle, John, 1965. What is a speech act? In Max Black (ed.), Philosophy in America
(Ithaca: Cornell University Press), 221-39.

Searle, John, 1969. Speech Acts (Cambridge: Cambridge University Press).

Searle, John, 1976. The classification of illocutionary acts. Language in Society, 5, 1-24.

Sudnow, David N., 1967. Passing On (Englewood Cliffs, NJ.: Prentice Hall).

Sudnow, David N., 1972a. Temporal parameters of interpersonal observation. In
Sudnow, 1972b, 259-79.

Sudnow, David N. (ed.), 1972b. Studies in Social Interaction (New York: Free Press).

Thibaut, John and Kelley, Harold, 1959. The Social Psychology of Groups (New York:
John Wiley).

van Dijk, Teun (ed.), 1985. Handbook of Discourse Analysis, volume 3: Discourse and
Dialogue (London: Academic Press).

von Neumann, John and Morgenstern, Oskar, 1944. Theory of Games and Economic
Behavior (Princeton: Princeton University Press).

Weber, Max, 1952. Ancient Judaism, translated and edited by Hans H. Gerth and Don
Martindale (Glencoe, III.: The Free Press).

Zimmerman, D. and West, C. (eds), 1980. Sociological Inquiry, 50, 3-4 [special issue on
language and social interaction].

Introduction

Beginning with the Fall Quarter of 1968, Harvey Sacks’ lectures were
delivered at the then recently established Irvine campus of the University of
California. The spirit of the new campus - at least of its School of Social
Science under the leadership of its Dean, James March - was quite in keeping
with the character of the 1960s. It was infused with a sense of possibility in
its academic and scientific ambitions and was correspondingly innovative in
organizational form. It dispensed with traditional academic disciplinary
boundaries and encouraged small groups of faculty to develop new research
enterprises and to define the terms - and requirements - of graduate degrees.
The central theme was the unleashing of high quality minds to follow their
scholarly and scientific instincts wherever the subject-matter, the theoretical
thrust, or the methodological possibilities seemed to lead, free of the
constraints imposed by traditional conceptions of disciplinary boundaries and
other “professional” obstacles to developments which could genuinely
surprise.

Whatever elements of his situation at UCLA suggested the possibility of
leaving, the animating ethos of Irvine’s School of Social Science was very well
suited indeed to Sacks’ own intellectual metier and character, and to the
disciplinary iconoclasm of his intellectual enterprise. It was a felicitous
matching of person and institution. Although Sacks developed a number of
close ties to faculty colleagues and played a distinctive role in the inescapable
politics of the academy - politics whose importance was amplified by the
minimized institutional apparatus of the School - in his work Sacks pursued
his own course and did not establish sustained collaborative undertakings
with others on the faculty. This too was a viable possibility within the School’s
culture. Sacks spent the remainder of his academic career at Irvine, although
at the very end he was considering another move.

There is little question that the character of Sacks’ work as it is displayed
in these lectures (as well as in those of Volume 1) was in various ways shaped

The introduction to Volume 1 presented some biographical information on Harvey Sacks’
education, and set the early phases of his work as presented in his lectures from 1964 to 1968
in the context of the academic social science of the time. That material is not repeated here,
and the reader interested in this background is referred to the prior volume. The present
introduction is concerned less with tracing linkages and contrasts between Sacks’ work and
other developments in social science (although there is some discussion of this sort) and more
with the treatment of Sacks’ work in its own terms.

I am indebted to Paul Drew and to John Heritage for reading a draft of this introduction
on my behalf, and for the collegiality and helpfulness of their responses.

X

Introduction

both by the larger social and cultural Zeitgeist of America in the 1960s 1 and
by the specific local ambience of southern California during that period,
within which the scene at Irvine played itself out. A delineation of those
connections will have to await another occasion. But there is equally little
question that Sacks’ oeuvre cannot be reduced to the socio-cultural environ-
ment in which it happened to emerge. The distinctiveness of his vision was
formed before the 1960s, and his pursuit of a distinctive path antedated that
special time as well. And it was formed not only in California but in such
bastions of academic tradition as Columbia, Harvard and Yale. 2

There is much continuity between the lectures published in Volumes 1 and
2. Most notably, the extraordinary, detailed analyses of small bits of
conversation in which whole social worlds and whole ranges of personal
experience are dissected from out of apparent interactional detritus continue
to be interlaced with more abstract theoretical and methodological discus-
sions. Various substantive themes persist as well - most importantly the
preoccupation with sequential analysis, and a continuing tacit preoccupation
with how to conceive of “culture.”

There are discontinuities as well. Topically, discussions of membership
categorization devices per se are not resumed, although on occasion the
resources of that body of work and the problems attendant on “doing
formulating” figure centrally, for example, in the lectures for Winter 1969.
A concern with storytelling in conversation which first emerges in the Spring
1968 term is much more fully developed, beginning with considerations of
sequential organization but extending into quite new analytic directions.
Observations about sound patterning and other “literary” aspects of word
selection emerge for the first time, and are taken up in several of the lecture
sets.

Thematic and analytic continuities and innovations aside, there are some
new stances taken up in the lectures published in Volume 2 to which it may
be useful to call attention, if only briefly. Some of these may serve to suggest
connections between the several sets of lectures which compose this volume;
others may serve as ways of focussing an initial orientation to each set of
lectures in turn. I begin with a theme which first appears in Fall 1968 but
recurs thereafter.

/

One apparent shift of stance which appears concomitant with the move to

'Recall, for example, (as a Los Angeles commentator recently did) that among the events
of just 1968 were counted “the year of McCarthy for President, the fall of L.B.J., the
assassinations of King and Robert Kennedy, the Beatles’ White Album, Motown and “2001:
A Space Odyssey”; of war, orgy and dreams of peace, in the summer after the Summer of
Love” (Los Angeles Times , August 23, 1991).

2 Views which reduce work like Sacks’ to something like the product of a California flower
child (for example, Gellner, 1975) are not only demeaning and intellectually evasive in
dismissing by epithet what they cannot decisively engage in substance; they are factually
ill-informed as well.

Introduction

xi

Irvine is a turn toward systematicity and toward the relevance of substantial
amounts of data, that is, aggregates of conversations or of instances of
particular phenomena in it. Consider, for example, the stance which Sacks
adopts in launching his discussion of turn-taking in lecture 3 (and continuing
into lecture 4) for Fall 1968. Among the key points in this new stance are the
following (all from lecture 3, p. 32):

What I want to do is to lay out in as general a way as possible at this
point how the sequential organization of conversation is constituted.

Note the shift to “general” and the generic reference to “conversation.”

I start out with two observations about single conversations. . .

Note that although the observations are about “single conversations,” they
are about aggregates of them.

I give in this first instance no materials for the observations, in that they
are grossly apparent.

Note this shift in practice; nothing in particular is the point of departure; an
observation about a regularity in an aggregate is the point of departure.

By the term ‘grossly’ I mean that while they’re overwhelmingly present
features, they are also sometimes not present features - and their
sometimes non-presence is something I will talk to at considerable
length.

Note that the issue here is the dealing with occurrences that depart from a
general practice, “sometimes non-presence.”

The shift, then, is to:

an order of organization , rather than a particular practice, of talking;

a class of places in an aggregate of data , rather than an excerpt;

an organizationally characterized problem or form of interactional work, rather

than an individually designed outcome;

invariancies of features rather than context-specified practices.

This is not, of course, a total shift of procedure. In lectures 5 and 6 Sacks
again presents particular materials, and explores turn-taking issues (among
others) in the context of a developing set of observations about that excerpt.
But there are readily observable consequences of this shift in point of
departure and analytic stance; one of these is an increasing (or increasingly
explicit) orientation to organization and structure in the domain of conver-
sational conduct. Again, discussion must be limited.

Near the beginning of lecture 4 (p. 44) Sacks develops the point that there

Introduction

xii

are grounds, built into the organization of conversation, for listening to every
utterance for any participant willing to speak if selected, or willing to wait if
another is selected to speak. The upshot:

So again there’s motivation to listen, which is independent of any rule
that would say ‘you ought to listen in conversation;’ motivation to listen
which turns on a willingness to speak or an interest in speaking.

This is not first time this point has been made in these lectures, but it has a
different resonance in the context of the new casting of turn-taking. What is
its interest?

Let us note first that the core point seems to be a grounding of listening to
utterances in the technical requirements of talking or not talking, the technical
requirements of the organization of proper conduct in that regard, rather than
its grounding in a normative injunction directed to that outcome specifically.
That is, an analytic concern for parsimony is at work here, setting aside a
normative constraint, a “rule” if you will, designed specifically to secure
“listening” or “attentiveness,” or showing that such a rule, if there is one, is
not there solely to secure attentiveness, because attentiveness is already a
natural, a technical, by-product of the organization of turn-taking . 3

Now this is surely not to deny a normative component to the organiza-
tion of interaction or conversation; surely, the sorts of mechanisms by which
the turn-taking organization is constituted are normative in character, for
the participants and consequently for analysts . 4 On the other hand, it does
seem to subordinate considerations which might be termed “politeness” to
ones which might be termed “technically constitutive” or “sequence
organizational.” The parsimony considerations here seem to take the form:
what sort of basic organization would both drive the prima facie organiza-
tion of the talk and engender whatever auxiliary effects seem to be
involved.

One implication is that listening is not vulnerable to (or is less vulnerable
to) whatever it is that may weaken persons’ commitment to observe

’Subsequently, in lecture 2 for Spring 1972, pp. 535-7, Sacks returns to this theme in the
context of his discussion of adjacency pairs, the virtually unrestricted freedom of occurrence of
their “first pair parts,” and the potential usability of first pair parts for selecting a next speaker.
From this, Sacks observes, it “falls out” that a participant willing to speak if selected to do
so will have to listen to everything said, for at any point a first pair part selecting them may
be done. This account in Spring 1972 is different only in its focus on adjacency pairs as
instruments of next speaker selection, rather than on the tum-taking organization per se.

4 And, indeed, at Fall 1968, lecture 4, p. 50, Sacks proposes, "We have in the first instance,
some formal normative features for conversation, which are in a way a public law for
conversation: One party at a time. . .” etc.

’Compare here the discussion (introduction to Volume 1, p. xxviii-xxix; 1 — li) of the
treatment of “indirect speech acts” with primary respect to considerations of politeness and
sequential organization respectively.

Introduction

xiii

normative constraints. 6 The point here is that the basis for listening is not as
much at risk as injunctions to be polite, when violations of politeness had
become, for example, a systematic political tactic on university campuses.
Listening was grounded in self-interest (wanting to talk, or being willing to)
and the technical requirements of implementing it. Departures from “features
of conversation” should be understood, therefore, not so much by reference to
motivated deviation from rules prescribing them as by reference to modified
operation of the system of which they are a by-product - for example, in
response to variations in context or transient problems in internal coordina-
tion.

Note the bearing of this tack on the claims of certain forms of “intention-
alist” theorizing (such as those of Searle, 1991) that our knowledge of human
action or conduct has only been advanced when “patterns” (as Searle calls
them) can be shown to be the causal products of intentions to produce them.
If the stance taken here by Sacks is correct, then observed distributions of
attention (i.e. , observed patterns of listening to others’ ongoing talk) may best
be understood not as the product of an intention to comply with a rule
mandating such attention (even if there was such a rule), but as an imposed
requirement for achieving such outcomes as talking if asked to, or withhold-
ing talk if another is asked to. (For a more general statement of this theme,
see lecture 2 for Spring 1970, and the discussion of that lecture below at
p. xxiv and n. 17).

Sacks’ grounding of the organization of attention/listening in the individ-
ual participant’s willingness to talk if asked to or to remain silent (even with
something to say) if another has been selected to talk, itself embodies a
distinctly sociological theme in accounts of social order. Developed in Sacks’
account of turn-taking most explicitly at lecture 4, pp. 50-2, this theme
understands the enforcement of the turn-taking organization to work by its
identification with individual participants’ rights and interests. So understood,
individuals are mobilized to defend their rights and interests (e.g., their turn
space); the emotions are recruited to this enterprise as well, such that
violations of “one-at-a-time” become treated as invasions of some speaker’s
right, and that incursion engenders anger in defense of those rights, that
emotional energy being put in the service of a socially organized enforcement
mechanism for the turn-taking organization. Further, gossip, reputation, and
the like can be recruited into that enforcement mechanism as well, e.g., under
the aegis of violators being “rude.” This, then, is how this class of violations
gets seen as violations of “politeness,” and it is in this light that we should
understand at least some “politeness” considerations. That is, it is by reference
to “politeness” that sanctioning is vernacularly formulated, while the actual
occasioning of the violations may be less a matter of normative etiquette
and its violations, and more a matter of technical organization or action
implementation, effectuated through the identification of individuals’

Something which was, of course, increasingly remarked upon in the 1960s, and certainly
not less in southern California than elsewhere.

XIV

Introduction

rights/interests with the resource which the turn-taking organization
distributes. 7

Throughout this discussion, it is apparent that considerations of systema-
ticity, structure and organization play an important role in understanding
orderly conduct observed across aggregates of data. 8 Although not all of the
Fall 1968 lectures display this stance, it does play a continuing (even an
increasing) role in Sacks’ subsequent work, including subsequent lectures, for
example, the lectures of Spring 1972 on adjacency pair organization.

II

Although “turn-taking organization” is the substantive focus for the Fall
1968 lectures, Sacks does not begin the course with a lecture on that topic;
indeed, he does not begin his discussion of it until lecture 3. The first two
lectures present another “take” on the “second stories” theme first treated in
the previous spring, at UCLA, and it may useful to linger for a moment on
what Sacks was doing in starting this course the way he did.

Note that the first lecture announces that it will be concerned with
something other than what Sacks otherwise plans to focus on. He begins:

Hereafter I’ll begin with some rather initial considerations about
sequencing in conversation. But this time I’m going to put us right into
the middle of things and pick a fragment that will introduce the range
of things I figure I can do.

He does this, he says, in order not to stake his claim on the usual insignia of
academic work (“ . . .its theoretical underpinnings, its hopes for the future, its
methodological elegance, its theoretical scope . . .”), but on the “interesting-
ness” of the findings. This was a task which Sacks set himself in the late
1960s - to have “bits” with which to tell lay people (including, for this

7 The theme of ensuring outcomes by identifying them with individuals’ property, interests
or rights - a familiar theme in certain '‘liberal” traditions of social theory - comes up again
in a strikingly different context in Sacks’ treatment of the motivated preservation of
experiences in memory for later retrieval and telling (cf. Spring 1970, lecture 5, pp. 257-9,
and below at pp. xxv-xxvi).

8 Another kind of consequence of this new stance, especially with respect to asserting claims
about aggregates of data rather than specific data fragments, is an occasional vulnerability in
the grounding of some claims in these lectures. Without materials as a shared point of
departure, it is at times unclear what actual things Sacks is talking about, and, therefore, how
to assess what he is saying. There are assertions, when the work takes this form, about things
which are said to happen "all the time,” which may not seem all that familiar to the reader.
(E.g., for this reader, p. 49: "Some people say about each other, ‘Why is it that we can never
have a conversation without it ending up in an argument?’ And in that it is a thing that is said
all the time, it is of interest to see how it could be sensible.”) Of course, what Sacks asserts
- at times ex cathedra - and the tack which he takes, regularly turn out to be of great interest
for their strategy of analysis even when subject to such reservations.

Introduction

xv

purpose, other “straight” academics) what “the work” consisted in which
would have a kind of transparent appeal and interest, readily presentable and
graspable in a relatively non-technical way, capturing “experiences” virtually
anyone would have had access to more or less directly, etc.

For a while, a regularly offered “for instance” was what Sacks proposed to
be an exemption from the ordinary recipient-design “rule” or “practice,” for
(among other forms of talk) storytelling - “Don’t tell others what you figure
they already know.” Sacks proposed that there is an exemption for spouses.
This is to be understood as a practice coordinate with a mandate to tell
spouses many things first, before they are told to anyone else. Then, given that
spouses are present together on many interactional occasions and that each
would have been first to be told most tellables, without the exemption many
tellables would have major constraints on their subsequent reliability to
others.

But the exemption engenders its own troubles. Because spouses’ presence
need not deter re-tellings, spouses may find themselves having to hear the
same stories over and over again. And the presence of an already “knowing”
person can have consequences for the form that the telling takes. As a result,
there is a pressure for the separation of spouses in social occasions where these
various cultural practices and orientations are in effect (thus, for example,
rendering them free for groupings based on other features, e.g., gender). 9

This was a neat little package, in which a familiar social experience did
seem readily traceable to practices of talking which ostensibly had little to
with them (or with anything of general interest), and was appealing and
satisfying as an “illustration” of the work.

Much in these first two lectures has the flavor that would make it attractive
on these grounds. Especially points well into the discussion of lecture 1,
regarding the counter-intuitive relative paucity of “things to talk about” with
those one has not talked to in a long time as compared with the ready supply
with those one talks to daily, 10 are just the sort of thing that Sacks saw as
useful in these ways. His departure from his planned theme in the initial
lectures in order to do this repeat “take” on second stories may embody his
treatment of the class members as part of a larger general public whch had to
be appealed to, at least initially, on the grounds of common experience. 1 1

One other aspect of these lectures which occurs in various of the sets but is
striking in the Fall 1968 set is what I will refer to as an aspect of their rhetoric.
One form which this rhetoric takes is the assertion, after some particular

9 A version of this line of analysis appears in this volume as lecture 4 for Fall 1971, where,
however, it is touched off by a particular data fragment, from which Sacks formulates the
problem of spouses’ talk.

ll> This theme is returned to in the initial lecture for Winter 1970, p. 172.

n In the lectures for Spring 1970, Sacks is explicit about the special cast being given the
first lecture. Strikingly, the topic which here in the Fall 1968 lectures serves as the accessible
beginning for the course becomes in the Spring 1970 lectures the “much more severely
technical” (Spring 1970, lecture 1) material which warrants a more accessible introductory
lecture!

XVI

Introduction

analysis or type of analysis has been offered, of its “normality” as a scientific or
disciplinary development. So, in Fall 1968, lecture 3, p. 38, Sacks proposes:

In its fashion the history I’ve recounted is a perfectly natural history; i.e.,
it would be perfectly natural for whatever course of development of
analysis of something that what you’re looking for initially when you
look at something - a plant, a social object, whatever it may be - is to
find some parts. One would begin off, then, with things like ‘greetings’
and in due course come to things like ‘one at a time’ and ‘speaker
change’ occurring.

Now, Sacks had read considerably in the history and philosophy of science, but
the claim made here is merely asserted and not developed by reference to that
literature. And what is asserted is an actual course of events of Sacks’ own
making, transformed into a putative generalized course of events which con-
stitute normality or “natural history.” The inter-convertability of modalities
such as instructions and historicized descriptions is something Sacks was well
aware of. It is a way of subsuming new departures, and a position staked out
without benefit of colleagues close by, under an umbrella of “normal science.”
Again at Fall 1968, lecture 4, pp. 54-5, Sacks invokes “naturalness”.
Having made a point about the co-occurrence of ‘one at a time’ and ‘speaker
change recurs’ as features of conversation that are “basic,” he then gives an
argument for this basic-ness (i.e., that the system is self-organizing, in that
breakdowns/violations are organized by reference not to some other rules but
by reference to these very same ones). 12 And then:

And I take it that that’s an extremely natural criterion for some rules
being basic; that is to say, when you reach them, you reach the ground.
There are no other rules which deal with how to deal with violations of
them.

It seems clear that this is not offered as an account of some actual history of
usages of “basic,” but as an effort to put into perspective the status of what
he was proposing. Here the rhetoric of “naturalness” is “aggressive,” in
claiming a status within some putative developmental course of a discipline.
Elsewhere, a more “defensive” (though hardly apologetic) tack is taken, as,

'“A similar argument is made with respect to adjacency pair organization in Spring 1972,
lecture 2; cf. below pp. xliv-xlv.

The contrast, it may be useful to mention (or one contrast at least) to this “self-organizing”
property is the sort of feature taken up in the 'Two preferences . . .’ paper (Sacks and
Schegloff, 1979), which is concerned with “second order organization.” There, if two features
meant to co-occur (in that context, “minimization” and “recipient design;” as here, “one at
a time” and “speaker change recurrence”) are not combinable on some occasion, there is an
extrinsic procedure for reconciling the conflict, i.e., relaxing one feature until the other can be
achieved. The parallel argument for “interruption” (as an instance of non-combinability of
“one at a time” and “speaker change”) being resolved in a “self-organizing” fashion has yet
to be presented formally.

Introduction

xvn

for example, in lecture 4 for Spring 1972, where Sacks offers an aside while
launching a discussion of adjacency pairs by formulating three abstract
utterance positions in conversation - “last, current, and next utterance.” He
says (pp. 554-5),

A lot of this will sound awfully banal but it’s far from that, so you’ll
have to jolt yourself - if I don’t jolt you - into thinking that it’s not,
after all, something anyone could have said; it’s not that it’s nothing; it’s
not that it has no consequences.

This should be appreciated as being at least as much self-directed as addressed
to the audience - either the physically co-present class or the audience
wherever. It is a sort of girding of loins before battle; a sort of assertion of
resoluteness.

In the intermissions and aftermaths of days we were working together,
Sacks used to bemoan the difficulty of the work. One of his metaphors for it
was the need to be able to “look around the corner of the future,” that is, to
be able to see ahead to that formulation of the organization of the world
which would appear in retrospect to have been obvious. And often this
seemed to turn on seeing in some (but not other) apparently commonsense
characterizations of empirical objects their potential for carrying heavy and
complex theoretic/analy tic loads. One problem which this posed was the
vulnerability to lapsing back into a mundane, vernacular, commonsense
hearing/understanding of those terms - one which would not sustain the
analytic load they were to carry, but would reduce to some “banal”
pre-theoretic assertion. It is that sort of vulnerability - both in his audience
and in himself, however differently for each - that this invocation seems
designed to confront; and it is similar vulnerability and transient self-doubts
which the “natural development” rhetoric seems designed to combat.

Ill

The Winter 1969 lectures presented here do not themselves compose a
thematically organized set, or even several such. Rather, they present a variety
of analytic topics and problems occasioned by efforts to come to terms with a
single stretch of material taken from the first of a series of group therapy
sessions with “adolescents” which Sacks had recorded (and, later in the
course, other materials as well). Although some considerations raised in
dealing with one part of this excerpt may come up in connection with another,
these lectures do not appear to have been designed to constitute coherent,
systematic treatments; still, in some instances (e.g., lecture 3) they do seem to
come together quite nicely. For the most part, however, some fragment of the
data segment is isolated for treatment, and then several sorts of interest in it
are extracted and addressed.

Not that this detracts from the striking and unexpected lines of analysis

xviii Introduction

which Sacks develops from his materials in the various, largely independent
discussions. The tone is set from the very beginning.

In the Spring 1966 lectures Sacks had examined the notion of “posses-
sion,” and in various respects reconstituted what sort of a cultural artifact it

is. In lecture 1 for Winter 1969 (although not explicitly related to the earlier
discussion), he makes another sort of novel use of “possession” or “owner-
ship.” In discussing the noticing/remarking by one participant in the group
therapy session on the hole in another’s shoe, Sacks notes that that the shoe
is owned by its wearer may entail that another cannot take it, but it does not
entail that another cannot talk about it. Further, if another talks about it, it
is very likely that its owner will talk next, or soon. So “ownership” is
conversationally consequential.

Furthermore, one of the generic matters conversation is centrally taken up
with is the things that the participants have brought with them to the
conversational occasion - their clothing, possessions, bodies, events they enact,
etc. The talk works off what the parties have brought; and parties can then
bring what they bring in part by virtue of the talk that may be made about

it. And persons may avoid being present to a conversation by virtue of what
they must necessarily bring to it (e.g., the current state of their bodies,
possessions, etc.), in view of the talk which that company is likely to make
about it. Possessions are then relevant not only to “the economy;” they are
central to the “conversational economy” as well. And “ownership” turns out
to be a social/sociological category which is consequential in hitherto
unappreciated respects.

There are other sociological threads running through many of the
discussions in these Winter 1969 lectures. One such theme concerns group
formation, membership claims, and different ways of “partitioning a
population” 13 to find who belongs together and who not. As the last of these
clauses may suggest, it is by way of interactants’ deployments of membership
categories and ways of identifying or formulating one another that these
various topics are addressed. In lecture 2, the issue is posed by how someone
is praised without impugning the status of the others (the issue being who is
the same category with the praised one and who not). In lecture 3 it is the
alternative ways of grouping two of the attendees of the therapy sessions -
Roger and A1 - together vis-a-vis the observer, as between patient/observer
and performer/audience. In lecture 7 it is the issue of who is a “hotrodder”

13 By “partitioning a population,” it may be recalled, Sacks refers to the results of
formulating a set of persons by reference to the categories in some empirically coherent set of
categories, i.e., categories which compose “a set” in an empirical sense. “Partitioning
constancy” (lecture 3, p. 110) describes the outcome when a same collection of persons are
distributed in the same way by reference to two or more different sets of categories. Thus, later
in this paragraph of the text, the category sets “patient/observer” and “performer/audience”
divide up the co-present persons in cognate fashion - the ones who are co-members of the
category “patient” in one set of categories being co-members of the category “performer” in
the other; these category-sets then display partitioning constancy for this population of
persons, or constitute “analog structures,” as Sacks also refers to the matter there (ibid).

Introduction

xix

(or “hippie”) and who not, who is “authorized” to make such a judgement
and how some persons “patrol the borders” of the category. In lecture 8, the
issue is posed by reference to alternative ways of seeing some collection of
persons in some place as legitimate or not, via their alternative formulations
as “gals and guys” or “den mother.”

Another, more methodological, theme which informs a number of the
lectures across considerable variation in substantive topic concerns the
relationship between “intuition” and “formal analysis” on the researcher’s
side on the one hand, and the relationship between analytical “formality”/
“abstractness” in contrast to the “concreteness” of “lived experience” for the
“ordinary actor” on the other.

Sacks’ characterization of what he is doing in lecture 2 - on “safe
compliments” - is instructive; its logic here echoes that of the analysis of
“invitations” as early as Spring 1966. In a discussion initially targetted at
“the weather” as a “safe” topic, Sacks begins elsewhere:

I did some work on ‘compliments,’ specifically on what I called ‘safe
compliments,’ the idea being to see what it was about some compli-
ments that made them ‘safe’ compliments, i.e., to turn an initial
observation into an analysis . . . The question then is, can we extract
from the sort of thing [some particular compliment} is, a set of features
which will locate a class of compliments like it, which are also safe
compliments? Where that is a test of the fact that we had some
generative features, [emphasis supplied}

Then, after developing an analysis of what makes one class of compliments
“safe.”:

Now the question is, with respect to ‘weather talk’, what do we need,
to be able to show that ‘the weather’ is a ‘safe topic’? What we need is
to develop a notion of ‘safe’ for topics so that we can have said something
when we say ‘weather’ is a safe topic.' The discussion on ‘safe
compliments’ was to give a sense that something could be done with a
notion of ‘safe’, something of a formal sort , i.e., it doesn’t have to be merely
an intuition , but what’s involved in something being ‘safe’ can be laid
out. [emphasis supplied}.

Now it should be clear from this treatment that what the professional analyst
might come to analyze as the formal features that make for “safeness” -
whether for compliments or for topics - is proposed to be “real” for parties
to talk-in-interaction; it is for them, after all, that it is proposed that the
“safety” matters, and they who may suffer from the lack of it. Still, such
formal accounts are vulnerable to charges of “formalism,” of imposing
analysts’ categories onto the lived experience of the participants, and the like.
To this theme it is useful to juxtapose Sacks’ discussion in lecture 3 of one way
in which two of the “patients” in the group therapy session deal with the fact

XX

Introduction

(of which they have been apprised) that there is an observer behind a one-way
mirror in the room. They “enact” a scene of “personnel just before a
performance,” calling out “Testing, one two three” and the like. It is in this
regard that Sacks points out the “partitioning constancy” in that setting
between “patient/observer” and “performer/audience,” which allows the
latter set of categories to provide a set of “cover” identities, at least
transiently. 14

Now this appears to ascribe to the teenage therapy patients a kind of
abstract or formal analysis of their circumstances which may appear to violate
our understanding of their lived experience. But Sacks argues (lecture 3,
pp. 1 10-11) that what is at issue in using a “theater” frame to deal with the
presence of an observer is that people

have their circumstances available to them in an abstract way, such that
they can use the abstract characteristics of their circumstances to locate
other circumstances that stand in a strong abstract relationship to their
current circumstances.

The relevance of this point is precisely to counter the objection to this whole
direction of analysis that, in explicating underlying abstract or formal features
of ordinary activities, violence is done to the lived-experience of those activities
for the actors who engage in them. By contrast, Sacks is proposing here that
part of ordinary Members’ competence is specifically an abstract understand-
ing of their circumstances and activities, an abstract knowledge drawn upon
in constructing further courses of action, and usable to construct further
courses of action in a fashion coordinated with others. Thus:

How can they use that abstract knowledge? They are able to use such
knowledge to locate circumstances which have features that stand in a
strong relationship to the initial circumstance, and those features are
then used to project actions by reference to those other circumstances,
which actions have some hope of being picked up. It’s not just one
person who is by himself capable of that, but he can have hopes that
others can see what he’s doing, see it fast, and collaborate with him.

The transformation by analysts of intuition into “something of a formal sort”
is thus not merely a requirement of disciplined inquiry, its results are
themselves meant to capture features of the procedures by which ordinary
conduct by ordinary members is methodically achieved.

14 See the earlier treatment of this episode in Volume 1, lecture 14 for Spring 1967.

Aside from the focus which the text brings to this discussion, Sacks’ demonstration of what
might be involved in seriously grappling with the effects which observers might have on a
“scene being observed’’ is a salutary one in refusing to settle for a simple and cliched concern
about “Heisenbergian” influences of observation itself. Rather, it insists on a detailed attention
to how and what sorts of changes in conduct there might be, how they are to be understood,
and how they would/might bear on what an observer makes of that conduct.

Introduction

IV

xxi

The lectures for Winter 1970 begin with a focus on the overall structural
organization of the unit “a single conversation,” linger in lecture 2 on the
theme of exploiting whatever topics come to notice in the intensive exami-
nation of a single conversation, and then return to considerations of overall
structural organization. There is much here that is penetrating and revelatory,
concerning such objects as ‘‘the reason for the call” and “reason for the call
relationships,” as well as “no reason for the call calls,” and relationships built
on them, to mention only some of the attractions of the first lecture.

There are elements in these lectures whose relevance is related to some of
the new emphases which I earlier suggested inform the lectures starting in late
1968. I want to take note in particular of a passage of two to three paragraphs
at pp. 168-9 of lecture 1 in which a theme first appearing in Fall 1968
reappears, and that is the relevance of examining a fragment from a
conversation in the context of (or juxtaposed with) other products of the sort
of “machinery” conjectured to be involved, other instances of the “same sort
of thing;” that is, the use of aggregates of data. In Fall 1968 this theme
surfaced in passing with respect to turn-taking; here it comes up in a more
sustained way with respect to the openings of conversations, both (and
especially here) on the telephone and in co-present interaction.

This is a topic - single case analysis versus working with collections of data
- which is not uncontroversial, and which Sacks and I discussed at
considerable length over the years. This is not the place for a thorough airing
of the issues or of those discussions. The key point here in Sacks’ treatment in
lectures 1 and 2, however, is that a proper grasp of what might be going on
in a conversational opening in some particular setting might require a grasp
of the range which the “machinery” involved in the production of the
phenomena involved could produce, and this might require examination of a
considerable array of data. 15

Once dealing with an array of data taken to be “comparable,” a
comparative analysis may appear to be needed, and this can itself give rise to
some methods of analysis which may obscure how the material being studied
may have been produced, rather than illuminating it. One such analytic
procedure requiring considerable care and reflection is “format-and-slot
analysis,” in which the prototypic problem is cast as a selection among
alternative terms which could be used for a same reference, or alternative
items which could be employed at a certain juncture in the talk, a juncture
formulated by the format of the talk in which it is embedded. It is not that
this form of analysis is flawed in principle; conversation analytic treatments of
reference - reference to persons, to places, etc. - have exploited it.

Sacks points out, however, that there are circumstances in which alterna-
tives to a term actually employed would/could not be used, even if they were
“correct. ” He takes as his case in point a telephone call in which the caller has

15 See the discussion below at pp. xxxix-xl and n. 28 .

xxii Introduction

called her friend about a commotion which was observed at the friend’s place
of work, a department store called “Bullock’s.” Sacks argues that in
proposing that one has called to tell another “what happened at Bullock’s
today,” the reference to “today” is not incidental. It is not properly
understood as being selected from a set of cognate temporal references. Had
the event happened several days earlier, the caller would not offer that
temporal referent in the same utterance format; she might not tell the story,
or find it tellable, at all. Indeed, given that the caller called in order to tell the
story, she might not have called at all. For it is its occurence “today,” Sacks
proposes, which makes the event “news,” and thus a possible “reason for the
call,” and hence in first topic position in the call. It is the fact that it was
“today” that makes for a temporal reference being used at all, rather than a
temporal reference being somehow slated to occur, with a selection procedure
then invoked to find the term to be plugged into that slot. And further, it is
not that its occurrence “today” makes it tellable as news per se; it makes it
tellable to one with whom the teller talks daily. It might not be tellable to a
twice yearly interlocutor, even if it happened “today,” for it may not have the
stature to be told in a six-monthly conversation. So all of the discussion is itself
subject to considerations of recipient design. These widening ripples of
analytic consideration surround the use of “slot-and-format” analysis, and
may render its invocation questionable.

Lecture 2 for Winter 1970 (at pp. 184-7) contains what is to my mind
one of the most striking discussions in all the lectures. Here Sacks turns a
seemingly technical dissection of the mundane story mentioned above - about
the commotion outside a department store told by one friend to another - into
a stunning demonstration of the alternative grasps of a scene which may
present themselves to different sorts of viewers - Sacks refers to it as having
become “kind of a distributional phenomenon.”

His account begins with the contrast between the actual teller’s perception
that there- was- trouble-and-the-police-were-taking-care-of-it on the one hand,
and, on the other, what Sacks proposes others (e.g., residents of the “ghetto”)
might see as there-being-trouble-and-the-police-were-engendering-it. He pro-
ceeds through a series of further related observations, for example, the
assuredness of the actual observer that her position as uninvolved witness is
unquestioned, as compared with the possibilities which other categories of
person finding themselves on such a scene would be required to entertain and
protect themselves against - for example, the possibility that they would be
treated as accomplices in whatever wrongdoing was suspected. The effect is to
render the scene which the story is intendedly about as equivocal as the
duck-rabbit of Gestalt psychology, and the actually told story as a situated,
perspectival version of it.

Sacks’ observations here carry the conversation-analytic treatment of an
ordinary story told in conversation to an intersection with traditional themes
of social and political analysis, and can well have served as a revelatory
component of a liberal arts education for white middle-class undergraduates
in Orange County, California in the aftermath of the Watts riots in Los

Introduction xxiii

Angeles just to the north, in years which were, in all but their numerical
depiction, still part of the 1960s.

This intersection with, and transformation of, vernacular understanding is,
I would like to stress, not a time-out from technical analysis but a product of
it. Sacks’ discussion here should be juxtaposed with his discussion of “viewer’s
maxims’’ in the paper ‘On the analyzability of stories by children’ (1972b) (or
the first lectures for Spring 1966 on which it was based), where the technical
basis for these observations may be seen to have been rooted.

V

The set of lectures for Spring 1970 is as coherent and stunning in its range and
perspicacity as anything in the collected lectures. It is the richest single set of
materials on Sacks’ treatment of storytelling in conversation, and surely
central to our understanding of stories more generally.

Here as before (cf. lectures 1 and 2 for Fall 1968) Sacks announces the
opening lecture 16 as one intended to appeal more broadly to the class than the
material to follow, which he characterizes as “much more severely technical
than most people could possibly be interested in.” He continues here the
practice of developing materials which could give “outsiders” a sense for this
work and its possible payoffs in a relatively vernacular way. The “more
accessible” materials of the Fall 1968 lectures, however, had become “much
more severely technical” by Spring 1970 (at least they were going to be
presented that way), and now were given their own, more readily accessible,
introduction.

Whatever the long term relationship of ethnomethodology and conversa-
tion analysis turns out to be, this lecture as much as any other in the corpus
of Sacks’ lectures (at least those to which we have continuing access)
exemplifies a convergence of the animating impulses of ethnomethodology
and conversation analysis in its preoccupation with the “ordinary,” the
“normal,” the “mundane” as achievements.

With lecture 2 Sacks begins the treatment of stories told in conversation.
It is a beautifully organized and accessible account of the sequential problem
of storytelling in conversation by reference to the organization of turn-taking
in conversation, and the understanding of the “story preface” by reference to
it (material later presented in Sacks, 1974). Perhaps two points may be
underscored here which might be overlooked in a reading of the lecture for
the aforementioned focus.

The first is Sacks’ self-conscious attention to theorizing as an activity. He
begins here - as he does in many other lectures - with what he calls an
“utterly bland fact,” one whose telling surely is not in itself of interest. The
point, he remarks, is what can be made of such a bland fact. But many bland

16 This lecture - supplemented by excerpts from lectures 2 and 4, and lecture 1 for Spring
1971 - has previously appeared in print (Atkinson and Heritage, 1984: 413-29).

XXIV

Introduction

facts lead to nothing beyond themselves. It is necessary, then, to have found
and pursued such a bland observation as allows something to be made of it.
In the end, then, the blandness or “obviousness” of some observation is
neither grounds for ignoring or suppressing it nor, in itself, for asserting it, but
for seeing if its achievement or consequences can be seen to be more telling
than the observation itself.

Such concerns with “theorizing” appear recurrently in this set of lectures
(as they do in the corpus as a whole). To cite but one additional instance, in
lecture 3 Sacks remarks on the common practice in everyday life that persons
take note of “coincidences” - for example, that they rarely go to some place,
and their interlocutor rarely does, and that it was a coincidence that they both
happened to do so on the same day and encountered one another there. He
then proposes:

I want to see if we can get at the beginning of an answer to how we come
to see these coincidences. The interest in the beginning of an answer is
not so much in whether it’s an answer - I don’t have any idea whether
it’s an answer - but in some way that the answer is built.

As with the blandness of the point of departure in lecture 2, the concern here
is with the ways of building an account, of theorizing in the presence of data
per se, rather than with the final assessment of the adequacy of the account.
By the end of the lecture, Sacks is again proposing that much of the
observable orderliness of the world may be better understood as the
by-products of ambient organizations which are quite unconcerned with these
outcomes, rather than as products which were the design target of some
organization. 1 7

In passing Sacks here produces an account of the perception of coincidences
that makes of it not a mistaken commonsense notion of probability, but
something like Marx’ notion of alienation; 18 that is, that persons’ own
activities (here the practices by which stories are formed up) produce a result
(an account of activities that is designed to make for relevant-at-that-moment
tellable stories), which is then perceived not as a product of the design of
storytelling, but as an independently encountered - and somewhat mysterious
- “external” reality.

Additional discussions of this explicitly methodological sort in the Spring
1970 lectures include an interest in “doing provings” (lecture 5, pp. 25 Iff),
“getting ... a problem” (lecture 6, pp. 267ff) and the relationship between
a “sophisticated lay observation” and more technical treatments (lecture 7,
pp. 271-2).

17 This is, then, a more general statement of a theme raised in Fall 1968, lecture 4, where
“listening in conversation” was treated as a technical requirement and result of the operation
of the turn-taking system, quite apart from any normative regulation explicitly concerned with
“listening in conversation.” Cf. that lecture, and the discussion above at pp. xii-xiv.

ls For example, the account of “alienated labour” in the Economic and Philosophical
Manuscripts (cf. McLellan, 1977: 77-87).

Introduction

XXV

A second point worth lingering on is Sacks’ treatment of the term “story.”
Especially in the years following these lectures there has been an explosive
growth in interest in, and writing about, stories, narrative, “narrative logic,”
etc., with whole fields and sub-fields (e.g., “narratology”) addressed to this
subject-matter. Unsurprisingly, the growth of this academic and literary
industry has spawned a profusion of definitions of the focal object — such as
“story.” Sacks parries the issues of “what is a story?” and “is this a story?” by
asking not whether the label “applies” (i.e. , is “correct”), but whether it is
relevant - that is, relevant to the participants in producing the stretch of talk
in and through which the object in question was produced. The issue is thus
transformed from an “external analyst’s” issue into a “a Member’s issue:”
how does it matter to the teller and the recipients that the talk being produced
(i in the course of producing it) is “a candidate story”? Lecture 3, and the other
lectures for the term, go to this question for stories in conversation in a fashion
that yields analytic leverage on the notion “story” for students of stories-in-
conversation distinct from stories in other contexts.

Lectures 4 and 5 present, respectively, an extraordinary discussion of
“entitlement to experience” (and to just the experience the events in question
will sustain) as well as of the cultural organization of experience and the
emotions, and a beautifully wrought account of “first” and “second” stories.
But what I would like to call special attention to is the way in which Sacks
brings an orientation to classical issues in social theory to a hypothetical -
but compellingly plausible - account of cognitive organization (lecture 5,
pp. 257-60).

Using the metaphor of “designing minds,” Sacks asks how the preserva-
tion of “experiences” might be organized. One cogent possibility might be to
store experiences by what would commonsensically be considered their most
important or salient aspects, or their most central character(s), or events, etc.
As an alternative he proposes the possibility that experiences be stored “in
terms of your place in them, without regard to whether you had an utterly
trivial or secondary or central place in them” (p. 258). The consequences
which this might engender - both for the organization of memory for
experience and for social intercourse about experience - are then cast in terms
of the concerns of social theory about the relationship between private interest
and the public good. 19

And that might have the virtue of providing a generalized motivation
for storing experiences. If it’s your part in it that you use to preserve it
by, then it might lead you to preserve lots of them, simply in terms of
the idea of experiences being treatable as your private property. People
can then collect a mass of private experiences that they then, by virtue
of their generalized orientation to ‘what’s mine,’ have an interest in
keeping. You might, then, design a collection of minds, each one storing

19 A theme which Sacks had invoked as well in the account of turn-taking presented in Fall,
1968. Lecture 4, pp. 50-4 and cf. discussion above, at pp. xiii-xiv.

XXVI

Introduction

experience which is to be used for each others’ benefit, though you
couldn’t necessarily say “Remember all these things so that you might
tell them to somebody else.’’ You have to have some basis for each
person storing some collection of stuff via some interest like ‘their own’
interest. Where, then, you get them to store experiences in terms of their
involvement, but have them be available to anybody who taps them
right.

This sort of linkage between social organization and the organization of
personal experience and its cognitive and emotional substrate - between the
social, the psychological and the biological - will surely have to be successfully
made eventually, and this is a novel and provocative direction in which it
might be pursued.

In its more immediate context, however, Sacks relates it stunningly to such
diverse ancillary themes as the personal experience of being understood or not
and the training requirements for professional therapists.

VI

Whereas the lectures for Spring 1970 were thematically coherent and
focussed on storytelling, the materials for Winter 1971 (very likely a graduate
seminar, rather than an undergraduate lecture course) deal with a congeries of
more loosely related matters. But the central preoccupation is with “word
selection’’ (cf. Sacks’ reference to “procedures whereby the words that people
use come to be selected;” March 4, p. 308), and in particular those
considerations of word selection that are often associated with “poetics. ” This
set of presentations (complemented by the lecture for May 17 in the Spring
1971 set, a lecture which deals with an eerie spatialization of metaphors,
idioms, and other aspects of the talk of both parties in an emergency
“psychiatric” phone call) constitute the basic point of depature in Sacks’
teaching oeuvre for this still largely unexplored domain of phenomena.

This central preoccupation aside, special attention may be called to the
presentation of March 1 1 which (at pp. 325-3 1) offers another one of Sacks’
astonishing tours deforce of analysis and interpretation. He starts with the text
of a sequence which seems to be ordinary enough, even if in it a couple appear
to press an offer of herring to an almost absurd extent. What Sacks does is to
lay bare layer after layer of organization and preoccupation (on the parti-
cipants’ parts) - from the differing grounds for making an offer than for
re-making it, to the tacit relationships between the parties that emerge into
relevance over the course of the sequence and come eventually to drive it, to
the ways in which processes such as those which this sequence embodies can
be a major component in both the stereotype and the enforced actuality of the
elderly in a society such as this - that is, the United States in 1971. We cannot
know whether the account which Sacks develops is biographically accurate for
this particular family, but it feels compellingly on target for the sorts of

Introduction xxvii

interactional processes which can constitute the lived interactional reality for
many persons. It is a signal display of Sacks’ ability to use a fragment of
interaction to capture in an analytically compelling way a whole complex of
social reality, from its social-organizational sources to its interactional embod-
iment to its experiential consequences. This discussion presents as well both
ends of a range of types of analysis which often appeal differentially to readers
of conversation-analytic work.

One end of the spectrum takes a particular episode as its virtually exclusive
focus, with its scope of generalization being defined by “however this analysis
turns out. ’ ’ Various particulars of context are traced through the full array of
their consequences; here, for example, that the offer-recipient - Max - is a
recent widower, and the offer-makers find themselves (on Sacks’ account)
newly responsible for his well-being. The contingencies of the offer and its
rejection, the relevance of pressing the offer and the import of its further
rejections - all are understood by reference to these attributes of the
participants, and the growing relevance of these attributes over the develop-
mental course of the sequence. The account thus appears compellingly
context-specific.

The contrasting end of the analytic spectrum focusses on the type of
sequence involved, across variations in particular settings of enactment. For
example, how is this sequence type — e.g., offers — related to other sequence
types? Sacks had a long-term ongoing inquiry on request and offer sequences,
and their relationship to each other and to other sequence types. Some of
Sacks’ students have also pursued these questions in this more categorical
fashion. Davidson (1984), for example, writes about “subsequent versions of
invitations, offers, requests and proposals dealing with potential or actual
rejection” (and see also Davidson, 1990). For dealing with the episode in the
March 1 1 session Sacks finds it more in point to juxtapose

an ‘offer’ as something different than a ‘request’ or a ‘warning’ or a
‘threat.’ But in some situations the offer is simply the first version of
getting the person to do something.

That is, the mode of analysis being pursued can lead to different sets of
alternatives providing the relevant comparisons or contrasts, “offer” making
such alternatives as “warning” and “threat” potentially relevant here, even if
they are not in other contexts.

Though there may appear to be a tension between these two modes of
proceeding, with the former often appearing more “humanistic,” “context-
sensitive,” and “holistic” and the latter appearing more “formalistic” and
“scientistic,” Sacks pursued them both. And the Winter 1971 materials show
them pursued hand in hand - the word-selectional considerations being
pressed in a generalized cross-context fashion, with this extraordinary single
case analysis occurring in the same class session.

xxviii Introduction

VII

In contrast to the Spring 1970 lectures which developed a coherently focussed
account of storytelling in conversation and did so by sustained examination of
a few data fragments, the Spring 1971 lectures (like those of the preceding
term) vary both in topical focus (from sound ordering to professional-client
interaction) and in data sources (from the “group therapy session’’ to a
telephone call invitation from one student to another to a call to a suicide
prevention center).

Among these lectures are included a series - those for April 30, May 3,
May 10, May 17, and May 21 - in which Sacks takes up the empirical
materials which he addressed at the start of the 1964-5 lectures. Readers
interested in the developmental course of Sacks’ lectures may wish to
juxtapose these two treatments, separated by some seven years of intense
intellectual work. It is not only that the same data are involved which might
inform such a juxtaposition, but that themes reappear in the Spring 1971
lectures which have not come up in the preceding several years. To cite but
one example, there is Sacks’ discussion (May 21, pp. 405ff.) of the
characterization of someone as a “stranger,’’ a discussion which goes back to
the issue of categorization (though not in that technical terminology) taken up
in detail in the paper “The search for help: no one to turn to” which was being
written just before and during the first of these academic terms of lectures.

Although the first lecture as delivered did include some initial discussion of
a data fragment, 20 it was largely given over to the stance which Sacks was
taking up with respect to his audience - both those present in the room and
those interested from afar (including, therefore, the present readership). It is
a rather franker statement than most instructors would give of the auspices
under which they address an undergraduate class. And it reverses the
relationship which might have been assumed to hold between the students
sitting in the room and those far away - in place or time - who might be
interested in “the work.” Rather than the latter being incidental and
“by-product” recipients of materials designed for the undergraduates, it is the
undergraduates who are recast as almost incidental onlookers to, and
overhearers of, this analytic undertaking.

“Almost;” for there is evidence throughout these lectures that the relevance
of the co-present audience did in fact enter into the shaping of the issues and
the manner of their presentation. There is, for example, the initial substantive
discussion. 21 Sacks explores some ways in which speakers find or select words

20 Cf. April 5 lecture, n. 1.

2 'As in the case of several previous consecutive terms of teaching, Sacks begins the
substance of the lecture set in the second of the consecutive terms with what he was exploring
less systematically in the preceding term. (See, for example, Winter and Spring 1970 on
storytelling in conversation.)

Note that parts of the text here have been rearranged for the sake of continuity and
coherence, so that some of the material included here with the lecture of April 5 was actually
part of the introductory lecture.

Introduction

XXIX

for use. In particular, he focusses on their doing so in “a history-sensitive”
manner, for example, by reference to the sound or (later on, in the lecture for
May 17) the metaphor composition of the prior talk. The tenor of the
discussion is instructive.

There are aspects of this discussion which suggest that, the stance taken in
the first lecture notwithstanding, Sacks did not entirely ignore the nature of
his co-present audience. The upshot which he takes from the discussion of
sound-patterning (lecture for April 5, pp. 341-4) is (p. 343):

for now . . . just to get some idea of how closely attentive in some
fashion people are to each other, where picking up the sounds, doing
simple contrasts, etc., are ways that they may be doing being attentive
to each other.

And again (p. 344):

when we begin to collect the sorts of things that I’m noting here, we can
feel that a serious attention to the way the talk is put together might
pay. These sorts of things at least suggest some sort of close develop-
ment.

And again, at the end of that lecture (in the present edition), after a discussion
of strategic considerations relevant to the parties in the talk in the group
therapy session materials (p. 347):

And that paralleling of the attention to a distinctive weakness can
suggest that they are moving with a kind of close attention to each other
in a conflictive way.

Two things may be said about the drawing of such conclusions. On the one
hand, they are in point for hearers with no previous exposure to conversational
materials and to this kind of close analysis of them. They seek to warrant the
kind of attention being paid to these materials in a way that would not appear
to be directed to an audience interested from afar in what Sacks has to say. On
the other hand, it was a task to which Sacks recurrently addressed himself -
to warrant these materials as respectable objects of study, and to establish over
and over again, in a variety of respects, that these materials were orderly at
quite refined levels of organizational detail. It is as if he were forever justifying

— to others and to himself — the undertaking, its starting point, and its key
premises. The upshots drawn here, early in the Spring term of 1971, can then
be understood to be addressed not only, or not especially, to the students in
the room, but to any recipients of his discussion.

There are two matters taken up in the Spring 1971 lectures which have a
history, either prior or subsequent, which it may be useful to call to attention

— the relevant identities of conversational participants, and the notion of
“preference.”

xxx Introduction

In the lecture for April 19, Sacks begins a discussion of “caller-called” by
reference to the possibility that

some part of a sequential organization of conversation has to do with
identities that the conversation itself makes relevant , such that for at least
those facets of the conversation one needn’t make reference to other sorts
of identities that parties have which are, so to speak, exterior to not
simply the conversation, but to its sequential organization. If, however,
we found that such other identities were central to almost anything one
could say about a conversation, then there would be a way in which
conversation could not be said to have an organization independent
from such other aspects of the world as yielded other identities, e.g.; the
names, sexes, social statuses, etc., of the parties. You could imagine a
world where some social status the parties had, operated in such a way
as to determine how 7 they could talk to each other, and in that world
conversation would not be an independently organized phenomenon.

The issue of the relevant formulation of the identities of participants comes up
recurrently throughout Sacks lectures. In the Spring 1967 lectures, it may be
recalled, there was a discussion of the possible “omnirelevance” of the
category-set “therapist-patient” for the group therapy session materials -
those categories straddling the line between “exterior” and conversation-
specific.

Elsewhere, in the lectures of Fall 1967 (and even more centrally in early
lectures for Spring 1967, not included in this edition) Sacks launched a
discussion of turn-taking by considering a claim in a paper by the anthropol-
ogist Ethel Albert (1964) about the practices of the Burundi. In this account,
members of that society are all hierarchically ordered, and the society is small
enough that on any occasion everyone present can assess their place relative to
everyone else present. The distribution of opportunities to talk is organized by
reference to this hierarchical ordering, 22 with the highest ranking person
speaking first, then the next, etc., until each has had an opportunity to speak
in an initial round; subsequent rounds reproduce this ordering.

Leaving aside a variety of problems which can be expected in a system
which worked in this way (and problems with the description), this account
embodies what Sacks has in mind by “a world where some social status the
parties had operated in such a way as to determine how they could talk to
each other, and in that world conversation would not be an independently
organized phenomenon.”

The point is that, if one could show for some culture / society that there is an
order or domain of conversation which is relatively autonomous of interac-
tionally extrinsic attributes, then the possibility of such a culture would have
been shown. Although it might be claimed in principle that there were other

22 As Albert put it (pp. 40-1), “The order in which individuals speak in a group is strictly
determined by seniority of rank.”

Introduction

XXXI

cultures where there was no such autonomous form, there would then be
certain burdens and opportunities of demonstration and exploration that
would have to be addressed. For one, it would have to be shown that there
was an empirical instance of such a society/culture, rather than simply be
asserted in the nature of the case. For another, it would be a feature of a
society/culture which might then be explored for what else it was related to,
how it came to be so, how it was embodied or implemented, etc.

This discussion, then, is intimately related to the issue of the omni-
relevance of gender or class status, etc. (class and gender are singled out here
because they are the features most often invoked as specially constraining and
shaping the conduct of talk in interaction). For to show a relatively
autonomous order of organization (or several such orders) for conversation
would be to establish domains of interaction not necessarily contingent on
gender, class, etc., and thereby to show conversation to be “an independently
organized phenomenon.”

One significance of the categorical identities “caller-called” is that they are
conversation-specific (unequivocally so, unlike “therapist-patient”), and it
appears that they serve as the feature by reference to which various aspects of
talk are organized, especially with respect to the overall structural organiza-
tion of single conversations. This had been shown in Schegloff (1967, 1968)
for the organization of openings, e.g., with respect to who talks first. Part of
Sacks’ argument here turns on the relevance of caller/called not only for
openings, but for “closings,” for example, it being the caller’s business to
initiate arrangement-making and other ways of getting to the end.

It is striking that in a prior discussion of omni-relevance (in Volume 1,
Spring 1967, lecture 14, and cf. the introduction to Volume 1, pp. liii-liv),
Sacks argued for the omni-relevance of “therapist-patient” in the group
therapy sessions by reference to its being the therapist’s business — in that
capacity - to bring the session to a close, and that a new patient has to be told
that an “indirect” closing initiation by the therapist was doing that job,
something which he did not himself see and which it would not have been
doing had anyone else said it. The relationship of some identity to a bearing
on “closing” (at least of a conversation as a whole) may, then, turn out to be
of strategic importance in showing category omni-relevance.

The issue of the relative autonomy of conversation/interaction has had a
continuing relevance for students of interaction. Perhaps the most prominent
discussion of the issue within contemporary sociology was Erving Goffman’s
presidential address to the American Sociological Association, ‘The interaction
order’ (Goffman, 1983) which also argued (albeit along different lines) for the
relative autonomy of the organization of interaction from other aspects of
social organization.

Another topic with a considerable later development figures in these
lectures for Spring 1971. At the end of the lecture for May 24 (pp. 414-15)
there is a discussion (the first of which there is a record, though Sacks refers
to an earlier related lecture) of the asymmetry of “yes” and “no” answers -
related to the form which a preceding “yes/no” question has taken. This is

xxxii Introduction

an early form of what would eventually become, under the name “the
preference for agreement” (cf. Sacks, 1987 [1973]), a much more general
account. Here Sacks appears to focus on such questions as might be termed
“pre”s, even if the future course of talk - the type of sequence - which the
speaker meant to undertake is not at all clearly projected. Whereas the
discussion in the May 24 lecture is quite specifically situated, and refers to
courses of action in which some sort of sequence may figure as preparation or
“setup,” Sacks would two years later, in the public lecture at the Linguistic
institute in Ann Arbor on which the 1987 publication is based, depict the
preference for agreement as a much more general - structural - feature of
question/answer sequences of the “yes/no” type, with still more general
implications, for example, for adjacency pairs.

A key component of this notion is that of “preference,” and it has a longer
(and variably focussed) history in Sacks’ oeuvre. In these lectures, for example,
in the lecture for April 23, Sacks proposes that some formulations of the event
for which an invitation is being tendered are “preferred:” if the occasion is to
include dinner, for instance, the invitation should be for “dinner;” anything
else (e.g., “drinks”) and recipients will hear that it is “not for dinner,” for,
given its “preferred” status as an invitation form, it would have been used if
it could have been used. And two years later at the Linguistic Institute, where
the “sequentialized” version of “preference” was extended from the usage
here to that of ‘On the preferences for agreement and contiguity. . . the
application of the notion “preference” to “formulations” was extended from
formulations-of-events to reference-to-persons in the drafting of the paper
‘Two preferences in the organization of reference to persons in conversation
and their interaction’ (eventually published as Sacks and Schegloff, 1979).

And before the usages here in the Spring 1971 lectures, a similar notion
underlies the conception of ‘specific alternatives” (cf. Schegloff and Sacks,
1973: 305, 313-14; this paper was first drafted during the summer of
1969). There the notion of “specific alternatives” made relevant by an
utterance such as a “possible pre-closing” was explicated by noting (p. 314)
that

the alternatives made relevant by an utterance of that form are not

symmetrical. Closing is the central possibility, further talk is alternative

to it; the reverse is not the case. . .

That feature of asymmetry - later central to the notion “preference” - came
up in other working sessions of 1969 and 1970. For example, I recall Sacks
remarking on it while examining tapes made by Melvin Pollner in a Southern
California traffic court; the observations concerned the treatment by the
parties involved of the source of income of a college student appearing before
the judge; Sacks took it that some sorts of financial support (I do not now
recall which) were central and “normative” (in the sociological terms of the
time), and others were alternatives to them, but not vice versa.

Introduction

xxxiii

The contrast figures, in essentially these terms, in lecture 6 for Fall 1971,
at pp. 455-6, where Sacks is discussing a story told by a teenage girl to a
teenage boy, a story which turns centrally on her spending half the night with
a “guy that {she} liked a real lot.” What is central to the telling of the story
is that they spent the time “in the back house” (i.e., the house behind the
main house, a sort of guest house) instead of “in a car.” Sacks shows how “in
the car” is built into the story as “normal” for teenagers, something with
which the teller is trying to fashion a contrast. “In the back house” is then a
specific alternative; it is an alternative to “parking” or “in the car,” but the
latter is not “an alternative;” it is the basic, unmarked (as linguists might put
it) place. And in that same context Sacks introduces the use of the term
‘ ‘preference:

. . .She can. . .invoke the normal priorities, in which, for unmarried
teenagers, parking is ‘preferred.’ I don’t mean that it’s favorite, but
there’s some way it’s preferred over the back house, if at least only in
moral terms. That is to say, she brings off that she prefers the back
house, but there is a more abstract sense of ‘prefer’ which involves her in
invoking the parking — that which is ‘preferred’ in the more abstract sense
— as a first alternative” {final emphasis supplied}

It is this sense of “preference,” as “a first alternative, to which others may
contrast but which itself does not contrast with them” which is one central
thrust of subsequent uses of the term, both by Sacks and by most others 23

VIII

Although the particular phenomena and data sources taken up in the lectures
for Fall 1971 are quite different, the thematic commitment underlying this
course is strikingly reminiscent of the lectures for Spring 1966. In both may
be found explorations of how (a) culture is to be conceived which blend a
fresh theoretical conception with a distinctive and organic relationship to
“ordinary” conversational data.

One relevant bit of background for the first lecture of the term may well
be an episode in law school (earlier recounted in the introduction to Volume
1, n. 6) which alerted Sacks to the mysteries of commonsense assessments of
the plausibility and seriousness of conjectured events. Law students debating
a point in the law of torts rejected as implausible the premise of an airplane
flying at an altitude of five feet while willingly discussing hypotheticals only

23 This includes, for example, Schegloff, Jefferson and Sacks, 1977. The notion captured by
“preference” figured in my own work at the time as well (especially in Schegloff, 1970), but
not under any of these names. For further discussion, cf. Schegloff, 1988, and, for another
view, Bilmes, 1988. For applications, discussions and reviews of the notion of “preference”
(and “dispreference”) cf. among others Atkinson and Drew, 1979: chapter 2; Heritage,
1984: 265-80; Levinson, 1983: 332-45; Pomerantz, 1978, 1984.

XXXIV

Introduction

marginally larger. The concerns then awakened, which had driven the
insistence on actual observation and then recorded and re-examinable data,
may have served tacitly as the grounds for the line developed at the start of
lecture 1 on the trouble of dealing with imagined occurences, and the
impossibility of dealing with events that strain commonsense credibility,
events which otherwise can be shown to be real.

This theme reappears explicitly later in the Fall 1971 lectures, in Lectures
9-12 “On some technical considerations of a dirty joke.’’ 24 There Sacks
points to a staggering number of implausible co-occurances on which the
story/dirty joke being examined depends, and without which it collapses.
And he addresses himself especially in lecture 10 to the devices by which the
telling of the joke/story can survive these apparent implausibilities. He rejects
an “Aristotelian” solution along the lines of a generic “suspension of
disbelief’ by noting that no disbelief arises to be suspended, and that the story
could not survive if it did. He suggests instead that a recipient is fully engaged
in understanding the story, and that the artfulness of the story in deploying
the elements from which an understanding can be achieved channels attention
in a fashion which circumvents the implausibility by naturalizing and
sequentializing the events.

Still, the isolation of this problem and the treatment of the narrative form
by ironic comparison to a quasi-realistic story suggests a continuing underly-
ing preoccupation on Sacks’ part with the relationship between the real and
the unreal, the plausible and the implausible, the real and the plausible, the
real and the implausible, etc. And here again (as in the Spring 1966 lectures;
cf. introduction to Volume 1, pp. xxxix-xli) may be found echoes of the
“commentator machine” introduced in Sacks’ early (1963) paper ‘Sociolog-
ical description,’ with its metaphorically articulated depiction of various
possible relationships between real doings and the accounts offered of them,
and the account-offering as itself a real doing relative to which another doing
may be a defective exemplar.

Lecture 1 begins an announced preoccupation with “storytelling in
conversation” with an observation about a pun, and the first several lectures
are as much about puns and proverbs as they are about storytelling.

Sacks’ concern with puns, which would eventually issue in a presentation
at the Georgetown Roundtable in March, 1972 and the little paper (Sacks,
1973) ‘On some puns, with some intimations,’ is analytically located at the
intersection of problems of word-selection of a “poetics ’’-like character on the
one hand, and the practices of storytelling sequences on the other. His
discussion of puns here in lecture 1 as well as in the Georgetown Roundtable
paper is focussed on their use 25 by a story-recipient just after story completion.
The occurrence of puns - unintended and unheard puns - in this distinctive
sequential position may have recommended itself to Sacks as a case in point

24 These lectures were published under that name, as edited by Gail Jefferson, cf.
Schenkein, 1978.

‘ 5 And the use of proverbs; cf. the ensuing discussion below.

Introduction

XXXV

for the contrast between “implausible ‘real things’ ” and imagined things one
could get someone to believe as a basis for theorizing about them. Or perhaps
the order was just the opposite; entertaining the possibility of opening the
lecture and the course with a discussion on puns, some groundwork seemed
called for, addressed discursively to the believability of the sort of thing he was
going to begin with.

The discussion of proverbs (at lecture 1, p. 422) goes back to Sacks’
reaction while still a graduate student at Berkeley to the beginning of George
Homans’ book, Social Behavior: Its Elementary Forms (1961). There (pp.
1-2) Homans remarks in passing on the traditional folk sociology which (it
was apparently his view) it was scientific sociology’s business and mandate to
correct and supplant. The hallmark of this folk knowledge was for Homans
apotheosized by proverbs with their “obvious” truths, but also directly
contradictory truths.

This theme was by no means unique to Homans. Both Homans and the
many others who contrasted scientific sociology with common knowledge
were engaged in defending sociology as an academic discipline from charges
that it was “nothing but common sense.” Many took it that one line of
defense was to show the weaknesses of commonsense knowledge, and thus the
proper office of sociology in reference to it. That office was to replace
“common sense” with something more scientific. This was, of course, one
central point of reference for Garfinkel’s observation (1967: chapter 1) that
the social sciences were addressed endlessly to the substitution of “objective”
for “indexical” assertions, and the alternative ethnomethodological program
which he put forward - to make commonsense knowledge non-competitively
a topic of sociological inquiry.

Sacks was struck early on (that would have been in the very early 1960s,
most likely in 1962-3, while we were at Berkeley’s Center for the Study of
Law and Society; cf. introduction to Volume 1, p. xv) by Homans’
non-analytical, non-sociological stance toward proverbs - treating them as
primitive and faulted versions of scientific propositions. The issue for Sacks
was, precisely, what were proverbs (as natural objects, so to speak), and what
were they used to do, that might make the features which Homans treated
ironically seriously understandable. He sought out a relevant literature and
found Archer Taylor’s The Proverb, but did not find the answer there, though
he respected it as a work of scholarship.

It is striking then to read Sacks’ treatment here (at lecture 1, p. 422)
with this history in mind. Briefly, in the context of a discussion of the use of
proverbs by story recipients on story completion, and having remarked on
the common observation of the inconsistency between different proverbs, he
asks,

Now the question is, is that a defect of proverbial expressions? Or is it
that, if it turns out that what proverbial expressions do is that they are
used to understand something else, then the question for them is, are
they applied to something that they evince an understanding of? If so,

XXXVI

Introduction

it’s quite irrelevant that, as a package, they can turn out to have an
inconsistency among them. The problem is not, on any given one’s use,
is it true relative to other proverbial expressions, but, does it, as
something one understands with, understand what it applies to? Where,
what it applies to is the story it’s used after. . .

. . .What’s done with them is to take one and see how, for what it’s
positioned after, does it understand that. It can then be seen as
irrelevant, somewhat arbitrary, to say “Let’s take the set of them and
consider whether they’re consistent, to determine whether they’re true.”
That may be not at all how, empirically, they work.

Here, some ten years later, is Sacks’ answer to Homans - his contrasting
account of how proverbs should be treated by sociologists. And in this little
passage is the direct confrontation of the effort to treat proverbs as defective
propositions - failures as “objective” expressions - with the claim that they
are designed fundamentally as objects for indexical deployment. They are
meant specifically to display understanding of the local object they are placed
after - they are prototypically indexical in that sense. Each is to be juxtaposed
to its occasion of use, for which it was employed; that specifically renders
problematic the detachment of each from the environment for which it was
produced, for juxtaposition with other such disengaged-from-context objects.
And Homans’ critique of them — based on just such a disengagement — is the
apotheosis of the social science practices to which Garfinkel meant to set
ethnomethodology in “non-ironic” contrast. For Sacks, this analysis grew
directly from his effort to figure out how proverbs worked.

IX

Those familiar with the published corpus of conversation-analytic work will
recognize in lecture 5 a version of the ‘Two preferences. . .’ paper (Sacks and
Schegloff, 1979). I do not recall at what point Sacks and I found ourselves
both focussing on the contrast between what Sacks here terms “Type 1” and
“Type 2” identifications of persons and what I was calling (with no restriction
to the description of persons) “description-for-recognition” and “description-
for-understanding.” The written version of the paper was initially drafted by
Sacks while we were living in the same house during the Linguistic Institute
at the University of Michigan in 1973. I did not know he was drafting it,
until he gave it to me early one evening to look over. Although we worked
over it intermittently, the changes made from the initial draft were relatively
small and technical.

The discussion of forms of reference in lecture 5 (as well as the paper which
followed) can be located in another course of development. While still at
UCLA (probably about 1966) Sacks had drafted a paper which came to be
referred to as the “two-person identification” paper. The data fragment which
had given rise to the “two-person identification” line of analysis, and around

Introduction

XXXVll

which the paper was written, was taken from the observational materials
collected by Barker and Wright. 26 In this episode, a little girl enters the
kitchen of her home and finds her mother talking to another woman,
someone the little girl does not know. The following exchange is then
reported:

Little girl : Who is she?

Mother : That’s Rita. Do you remember the other day when you went

to the party and met Una? Well that’s Una’s mother.

This data fragment was appreciated against the following analytical back-
ground.

Sacks had established in his dissertation work (cf. the published version in
Sacks, 1972a) that there was no general solution to what he termed “the
one-person identification problem.’’ That is, faced with the task of
identifying/categorizing a single person, there were demonstrably available
multiple “membership categorization devices” which contained some cate-
gory which could properly categorize any person. 27 And there did not appear
to be any general solution for selecting which device to use - no general
preference rule that would select some device from among whose categories
“the correct one” for the person being categorized should be selected. This
was a finding with many analytic and theoretic reverberations. For example,
analytically, any actually employed categorization employed by a speaker in
talk-in-interation had then to be viewed as a contingent product whose
achievement could be subjected to analysis by reference to the particulars of
its local environment. (And, theoretically, social scientists’ categorizations
could not be warranted solely on the basis of their descriptive correctness, but
had to be otherwise warranted, e.g., by reference to their relevance, whatever
grounds of relevance might be chosen.)

What the data fragment reproduced above suggested to Sacks was that,
whereas there seemed to be no general solution to the owe-person indentifi-
cation problem, there might be a solution to a two - person identification
problem. In his dissertation (1972a), he had described what he proposed to
be a categorization device composed of pairs of linked terms - “paired
relational categories” he called them - (e.g., friend-friend, husband-wife,
relative-relative, parent-child, neighbor-neighbor, stranger-stranger, etc.),
which constituted “. . .a locus for a set of rights and obligations. . .” (p. 37).
This categorization device was used to categorize a population of persons not
one at a time, but two at a time — as incumbents of one of these paired

26 Although Sacks had worked on some observational materials which Barker and Wright
had published (for example, One Boy's Day , 1951), I believe the fragment involved in the
“two-person identification” paper was taken from other, unpublished, material of theirs
which Sacks had secured.

27 Sacks had termed these devices “Pn-adequate,” i.e., adequate for any, unspecified (hence
“n”) population (hence “P”). The devices/collections of “age” and “sex/gender” categories
were his most commonly invoked instances.

xxxviii Introduction

relational categories. It appeared as well that there was only one such
categorization device, only one which identified / categorized persons two at a
time.

What the “Rita/Una’s mother” data fragment suggested was that one
way members might handle a one-person identification problem which had
no general solution was to transform it into a two-person identification
problem which did have a general solution. In the instance at hand, the little
girl’s mother adopts this solution: asked to identify one person (“Who is
she?”), she introduces another person into the identification problem - Una -
and then identifies the pair of persons by a set of paired relational categories:
mother-fchild] (“. . .That’s Una’s mother”). This was an extremely elegant
solution to the identification problem, and an extremely elegant account of it.

But there were problems, and on a visit to the west coast during the winter
break Sacks and I discussed them at length, as we regularly did with one
another’s written work. The most telling - and ultimately fatal - problem
was that this solution did not work as a general solution. For one thing, not
all the paired sets of terms could be (or were actually) used by interactants; for
example, although “stranger-stranger” was one of the paired term-sets (and
one indispensable for the empirical context which first gave rise to the
formulation of this categorization device), persons confronted with an
identification problem do not respond by saying, “That’s Rita. Remember
Una? Una and Rita are strangers.” Were stranger-stranger an eligible
category-set for these purposes, there might be a general solution to the
one-person problem by converting it into a two-person problem. Without it,
it was not a general solution.

Another problem, equally fatal and with clear connections to the lecture
which prompts this discussion (and to the ‘Two preferences. . .’ paper), was
that not any person could be introduced as the second for co-categorization
with the initial person to be identified, and not even any person in a specified
range of relationships to the target problem. Only such persons could be
introduced (or seemed actually to be introduced) as were expectably recog-
nizable to the one posing the problem, the one for whom the categorization
was being done. So again, persons confronted with an identification problem
do not say “That’s Rita; there’s a little girl named Una, and Rita is Una’s
mother.” There was, then, not only a constraint on which set of paired terms
could be used for the target person and the one to be introduced as second;
there were restrictions on which second person could be introduced for this
purpose by reference to the knowledge of the recipient of the identification.
Indeed, the possibility could not be ruled out that no second person could be
found who would satisfy both containts (nor was it clear that these were the
only constraints). The status of this categorization device as a general
solution to a two-person identification problem was thus cast into doubt, let
alone its status as second-order solution to the one-person identification
problem. The “two-person identification” paper was shelved. But it was
not without consequences, of which brief mention can be made here of only
three.

Introduction

XXXIX

First, the data fragment which motivated that earlier effort has here, in the
Fall 1971 lectures (lecture 5, pp. 451-2) become an example of

where a speaker doesn’t figure that recipient knows who’s being referred
to, but knows something that involves it in being an ‘almost,’ i.e., that
you know someone in some close relationship to that one being referred
to.

Its bearing thus is incorporated into the discussion of “recognitional refer-
ence” and the preference for recognitional reference even when the possibility
of its achievement is open to question (cf. Sacks and Schegloff, 1979).

Second, it seems to be relevant in a curious way to a tack taken in an earlier
lecture, and on quite a different topic - lecture 6 for Fall 1968. A bit into that
lecture (at p. 70), Sacks is discussing introductions (of one person to another),
what occasions them and how they’re done.

One way to think about it is to consider that a way to simplify the task
of doing any introduction would be, e.g., to constrain the occasions
under which introductions could get done. You could say, for example,
introductions should go ‘first name to first name.’ That can operate to
constrain the initial use of an introduction to only people you can
introduce that way.

But, Sacks points out, that runs up against the fact that the conversations
within which introductions have to get made are generated by an entirely
separate mechanism from the one that makes introductions possibly rele-
vant.

The relationship to the problems with the “two-person identification”
paper is this: one problem with that paper, as just recounted, was that the
mechanism only worked for certain possible values of paired relational
category terms (not for e.g., stranger-stranger), and setting such a pre-
specification subverted the potential generality of the device. So here as well,
where the point is that an introduction mechanism is needed which will have
as general a scope as whatever occasions the relevance of an introduction and
whatever occasions the already-ongoing conversation within which introduc-
tions come to be relevant. Pre-constraining introductions to certain values of
introduction terms would subvert the viability of that institution. This is just
another specification of the more general result that pre-constraining the
elements of a device which can be employed subverts the possible use of the
device as a general solution to some problem in the practices of interaction.

Third, the working through of the problems of the “two-person identifi-
cation” paper seems to have deeply affected Sacks’ thinking about the relative
merits of single case analysis versus the use of aggregates of data for the
purposes of building a discipline. Note that the issue is not the status of single
case analysis per se, but the possibility of building the sort of desired discipline
which had come to be the goal of conversation-analytic work. In a letter to me

xl

Introduction

a few years later (March, 1974), Sacks remarked on the relevance of
“working with masses of data” as what “in the end differentiates what we do
from e.g. French structuralism.” And, in this regard, he invoked the
experience of the problems with the “two-person identification” paper - and
its effort to ground a general solution in a single case - as evidence enough. 28

X

Having initially projected the Fall 1971 lectures to be about stories and
storytelling, the first six lectures depart somewhat from a close focus on that
topic, although remaining at least tangentially relevant with lectures 7 and 8,
and then with the series of lectures from 9 through 12, Sacks comes back
squarely to his announced topic.

Lectures 7 and 8 address the “motive power of stories.” The theme is a
penetrating and remarkable account of a particular class of stories. These are
stories which come to be retold after a long time delay (“long” here meaning
years), a delay during which one who had been the recipient of the story
becomes the kind of person the teller then was, and tells it in turn to a
recipient such as he was when he was told it - the retelling being done on just
the sort of occasion which is appropriately analyzed by the story. Such a
“delayed-fuse” story thus serves as a kind of cultural repository for
occasion-ally relevant knowledge. (The material being analyzed involves an
older man, seeable as “no longer having prospects,” telling a younger man,
who is about to depart for college - and prospects - about the time he was a
young man, with prospects, and what became of them.)

These lectures call to mind the lectures of Spring 1966, for the way in
which they speak to the nature of culture, the ways in which culture mobilizes
minds as a repository of what it has to transmit, and uses stories as the vehicle
for transmitting that knowledge, recruiting the interactional stances of the
participants in the situations in which they find themselves - for which the
stories provide analyses - as the energy driving the telling of the stories as
matters of e.g., self-justification. They also recall lecture 5 of Spring 1970 on
how memory for experiences can be motivated by having them stored as “the
property” of the one to remember them, to be accessed by others by telling
a “similar” story.

The theme plays off a by-now cliched geneticist “witticism” that chickens
are the device by which eggs reproduce themselves. Here persons, their
experiences, and the stored versions of experiences in stories are the device by
which culture reproduces itself and adapts to changing social circumstances.
The line taken here is reminiscent of a term (though not necessarily the

28 He wrote, “The ‘structures for particulars’ direction [which is how Sacks had earlier
characterized “the thrust of my stuff over the years”] doesn’t work: recall the two-person
paper failures, etc. and the ‘system for masses,’ for routine, etc. may.” (The internal quotation
marks have been added for clarity.)

Introduction

xli

correlative meaning) which the anthropologist Alfred Kroeber introduced
some years ago for culture - “the superorganic.” 29

These resonances of lectures 7 and 8 are sustained in the following four
lectures, 9 through 12, on the dirty joke as a technical object. In this
discussion as well, the story form is treated as a packaging device for elements
of culture, as was the case in lectures 7 and 8. There is a distinct shift in theme
here, focussing less on the teller doing things via the story and more on the
story doing things through the teller, and doing them through the teller as the
instrument of a culture. The story in general, and the dirty joke as a technical
object in particular, get worked up somewhat formally here in a fashion
parallel to the account of games (and children’s games in particular) as
packaging units for a culture in the Spring 1966 lectures.

This is a weighty theme and it may be appropriate to understand Sacks to
have prepared his audience for it in the opening lectures for the term. Recall
that in the first lecture in this set for Spring 1971 Sacks had tried to provide
grounds for taking seriously the possibility that there really was a pun in the
story, that it was not just a “reading-in” by the analyst, just as he had done
in other first lectures, to ground the seriousness of word-selectional or
“poetics” observations. Here he proceeds in the same fashion by showing the
“artfulness” of the dirty joke/story, the elaborate way it is put together in
order to ground a claim for its status as a technical object, and eventually his
claim for it as a serious transmitter of culture.

It is in lecture 1 1 that the theme of the dirty joke as a packaging device for
culture, with its “dirtyness’ serving as a form of restriction on its circulation,
is stated most pointedly. It may be worth mentioning here again (cf.
introduction to Volume 1, p. xxii) the relevance of the work of classical
scholars such as Milman Parry, Albert Lord, and Eric Havelock, all of whose
work Sacks was familiar with, and from whom he would have become
familiar with the notion that the classic forms of oral cultures - such as the
Greek epic - served as major instruments for the preservation and transmis-
sion of a culture, the story line of the epic being not so much the point of it
as the shaper and guarantor of its transmission. It was just one aspect of the
special metier of Sacks’ mind and sensibility to see in this juvenile “dirty joke”
told in a teenagers’ group therapy session the contemporary operation of so
grand a theme, otherwise treated as the special preserve of elite “culture.”

Another echo of the Spring 1966 lectures in Fall 1971 is the appearance
of a concern with children, and children’s learning the ways of the culture and
its rules, a theme which is central here in lecture 12. This lecture again calls
to attention Sacks’ extraordinary capacity to take apparently general views
and characterizations of the world, ones which present themselves as
“natural” accounts of it, and to specify them, often showing them to embody
some distinct and limited perspective. Thus in lecture 12 he depicts what
seems to be a potentially anybody’s recounting of a scene as specifically
embodying the perspective of 12-year-old girls. In the earlier lectures 7 and

29 Cf. Kroeber, 1917.

xlii

Introduction

8 for Fall 1971, he shows how a story embodies in particular the perspective
of persons with live prospects for a future or those with already failed
prospects. And in an earlier set of lectures (Winter 1970, lecture 2) he takes
what appears to be a passing observer’s “bland” general account of a scene in
which “the police were handling some trouble at a department store” and
shows that to other eyes - of members of different social groups with a
different experience with authority - what might be seen might be not that
there was trouble and the police came and handled it, but that the police came
and there was trouble and it was unclear how it was being handled, how it
would turn out, and how it would turn out to implicate them.

Although seemingly quite remote from that tradition of analysis, these are
exemplary exercises in the sociology of knowledge. Apparently unsituated
views and understandings of the world and of particular settings — otherwise
understandable as just “how things are/were” - are analyzed for the
distinctive social groups to which they are affiliated, and with whose
experience of the world they link up. These discussions illuminate our
understanding both of the particular settings and utterances being addressed,
and of the distinctive experience of social circles to which we gain access by
way of these discursive practices.

In this regard, it is striking that one of Sacks’ characteizations of the special
perspective of 12-year-old girls by reference to which the dirty joke being
examined in lectures 9-12 should be understood is reminiscent of his
depiction of the perspective of suicidal persons who see themselves (and report
themselves) as having “no one to turn to.” That phrase supplied the subtitle
of Sacks’ major early paper (and his dissertation), The Search for Help: no one
to turn to. Now remembered primarily for its formal statement of the
categorization problem and aspects of its solution, it may be useful to recall
that, although textually at the beginning of that work, developmentally it was
subsequent to the initiating problem, which was how someone might come to
say “I have no one to turn to,” and say it seriously (that is, as the reported
result of a search procedure), delivered paradoxically precisely in an occasion
in which it seems apparent that they have found “someone to turn to.” Sacks
began with that, although in the paper he ends with it.

The proximate solution of “no one to turn to,” Sacks proposed, was that
the person involved (the suicidal person, that is) had such a problem as would
alienate precisely the person(s) whom the normative search procedure would
locate as the proper persons to turn to. That is, there are in general “persons
to turn to” (formulated by reference to paired relational terms discussed
above at pp. xxxvii), but the problem involved, if reported to those persons,
might lead to their abandonment of just the status which made them the
one(s) to turn to. Thus, for example, turning to a spouse with a problem
engendered by one’s adulterous involvement.

What is striking is the formal similarity to this situation of the putative
circumstances of 12-year-old girls in Sacks’ account of the dirty joke: namely,
the problem of checking out information about sex, information acquired
illicitly, e.g., by listening in to the parent’s bedroom from behind a door: with

Introduction

xliii

whom can such supposition be checked? the parents one spied on? the friends
to whom one cannot reveal just this inexperience? The formal similarity is
striking: the nature of the problem is what precludes turning to just the ones
one would otherwise turn to for its solution.

The last lecture for Fall 1971 is about dreams, and seems quite disengaged
from the other lectures. In fact, Sacks developed a considerable interest in
dreams (among other respects as a format in which stories are preserved), and
pursued it, largely informally, during these years. In part, this had developed
from his reading in Freud and in a variety of literary sources; in part it
converged with an interest in popular culture (an interest which, in the
last several years before his death, included such matters as advertising as
well.) He was, for example, interested in the presentational modality of
dreams - whether they were experienced as being read or being seen in action;
if seen, like a movie, whether they were in color or black and white; what
sorts of editing and directorial techniques informed their structure, and the
like.

XI

The lectures for Spring 1972 begin in the same fashion as did those for Fall
1968, as a systematic and general account of an organization or a class of
conversational occurrences - in this case, adjacency pairs and adjacency pair
organization. It is not until the second half of the second lecture that a
determinate, actual (as compared to intendedly exemplary or “characteristic”)
bit of talk is presented for careful and detailed examination. But the text of
the first lecture and a half nonetheless makes clear that this general and
systematic introduction to the projected subject-matter for the course is based
in a detailed way on a substantial corpus of observations and analyses of
particular stretches of talk of which adjacency pair organization is to be offered
as a tentative account (though hardly preparing us for the illuminating detail
exposed when the first bit of data is examined closely in the second half of the
second lecture).

The general features of adjacency pairs are first introduced via a variety of
particular sequence types - greetings, terminal exchanges, question-answer
sequences, etc. - each of which names its own, recognizable class of sequences.
Adjacency pairs are thus introduced as a class of classes. But the particular
variety of sequence types is strategically selected to display something of the
extraordinary provenance of adjacency pairs - used at the critical junctures of
virtually all the main kinds of organization of conversation: at the opening
and closing boundaries of particular episodes of conversation, as the central
device by which next speakers are selected, as the basic tool for remedying
various locally occuring problems in conversation, as the locus for departures
from a single-sentence format for utterances (sub-sentential utterances char-
acteristically being second-pair parts, and the construction of multi-sentence

xliv Introduction

utterances being mediated on this account by adjacency pair constructions),
and so forth.

This introductory account of the generality of adjacency pairs by reference
to the other types of organization in which they figure prominently and
strategically is followed by another, an account of their provenance by
reference to their distributional generality. That is, if we ask where adjacency
pairs can go (and, in particular, where their first pair parts can go, since where
the second pair parts go is given by the first pair parts, i.e., after them), we
find that their privilege of occurence is unrestricted except by reference to
adjacency pair organization. That is, they can go anywhere except after a first
pair part, unless the one going “after” is initiating an “insertion sequence.”
The point here is two-fold: our sense of the centrality of adjacency pair
organization is reinforced by its virtually unrestricted distribution, and our
sense of its basicness is reinforced by its self-organizing character, that is, by
the observation that the only restriction on its distribution is that imposed by
adjacency pair organization itself. (Recall that a similar argument had been
offered for the basicness of the turn-taking organization in Fall 1968, lecture
4, pp. 54-5, and this introduction, above, p. xvi and n. 12).

When Sacks turns to the examination of a specimen of an adjacency pair,
the focus shifts sharply. The exchange - a question/answer sequence -
occurs in

a telephone conversation between two middle-aged women one of
whom has gone back to college part-time, and is telling the other about
a class she’s taking

The other - Emma - asks:

Emma : Are you the oldest one in the class?

Bernice : Oh, by far.

In some five pages, Sacks shows an array of issues to be involved which most
readers, I suspect, will not have anticipated. Here I want to draw out one of
them, one which echoes themes raised in earlier lecture sets, especially that for
Spring 1966 (and see the introduction to Volume 1, pp. xxxvii-xxxix). The
issue concerns the proper understanding of the positioning of the subject-
matter of these lectures - and of the area of inquiry which has developed with
the name “conversation analysis” - among the disciplines.

One of Sacks’ early observations about this exchange is that the question is
not characterizing Bernice’s position in the class as one of a possible set of
positions, others of which might be “second oldest,” or “one of the oldest,”
and the answer is not just a way of saying “yes,” or saying it emphatically.
Rather, Sacks proposes with respect to the former, the question is asking
about a “unique position” in the class, with a variety of features which can go
with occupying a unique position (“being the only X”); in that respect, its
relevant alternatives are not the set of age-grade positions, but things like

Introduction xlv

“Are you the only woman? Disabled person? African American?” etc. Sacks
continues (lecture 2, p. 538),

So that what seems like a kind of obvious semantics turns out to be
wrong for our language. It’s one you hear around, and it says: Take “the
oldest one in the class” and find its meaning by considering the set of
alternatives to it, where the alternatives can easily be derived from it by
just considering some obvious way in which it is part of a set of positions
having to do with ‘oldness.’ . . . Now, alternatives are an obvious way
to go about locating what something is doing or what something means.
But the question of alternatives does not have an easy answer. It is, for
any given thing, an empirical issue and not simply a transparent
semantic issue to be gotten by lexical considerations. In saying what I
figure to be the kinds of things that are alternatives here, both in the
question and in the answer, I’m saying something that has to he
discovered from a consideration of the way the world works that produces
these kinds of sequences. This obviously produces a massively complex
set of problems in analyzing things like a small question-answer
sequence. For each one of them, if we’re going to use alternatives to find
out what it means, then we’re going to have to go into a discovery of
what the alternatives are. [Second emphasis supplied}

The point to which I wish to call attention is that this is not a matter of
linguistic analysis in the usual sense; the closest might be some form of
anthropological linguistics or linguistic anthropology, though those disciplines
have shown qualified enthusiasm at best for this sort of analysis. The point
here echoes a point like the one made in the Spring 19 66 lectures apropos of
“possessive pronouns;” they work linguistically as possessives only given an
independent analysis of what they are affiliated to as “possessable” (hence the
very different senses of “my shoes” and “my barber”). And the latter are not
linguistic facts. 30

But more is involved than there simply being a separate domain to be
studied here, and therefore possibly a different discipline. When he turns to
Bernice’s answer, Sacks notes that it says “The question you asked me is
correct. I am what you’re supposing I am.” Then (p. 539):

And by using “by far” one indicates how one would know it; i.e., by
looking around the class, without any particular interest in finding out
the ages, she could age herself relatively to everyone else - which is after
all not a thing that many in a class would do. But there are some people
who can do it just like that, by virtue of that it’s a ‘by far.’ That is to
say, ‘by far’ is glance-determinable . And if it’s glance-determinable, then
that’s how you could have known it . . . It’s visible, like anything else

30 This discussion is clearly related closely to the one about “frame-and-slot” analysis in
lecture 1 of the Winter 1970 set, and cf. above at pp. xxi-xxii.

xlvi

Introduction

in the room, that she is older by far. And as she knows it, so does

anybody else in the class know it.

Sacks then points out that “that the answer says how one knows what one is
saying is a common feature of answers.” This is the sort of thing that linguists
(e.g., Chafe and Nichols, 1986) mean by the term “evidentials.” But “by
far” is not, I believe, the sort of item (such as modals like “must have,”
attributions like “John said . . . ,” access routes such as “I read that ...” etc.)
that is ordinarily counted as an evidential. It is not a linguistic feature, but a
grasp of the course of action by which such a formulation would come to be
made, and via an appreciation of its consequentiality to the circumstances of
the one making it, that “by far the oldest” as glance-determinable needs to
be understood. For while “by far” may have these attributes for this question
by this asker to this recipient about this setting, it is by no means a feature of
its linguistic realization per se, or even one of its variants. The range of further
observations which an exchange like this can engender, and the theoretical
directions in which they lead (both of which Sacks pursues in the remainder
of this discussion) belong to a domain of inquiry that may well be a necessary
complement to a thoroughgoing linguistics but is not part of it, and should
be part of a thoroughgoing sociology or anthropology, but does not seem
likely to become that either.

The Spring 1972 lectures present various of the juxtapositions or contrasts
which run through Sacks’ oeuvre. Lectures 1 and 2 juxtapose discussions of the
most abstract and general sort - characterizing a formal structure, the
adjacency pair, not only as a type or class of occurence, but as a class of classes
- with a detailed examination of a single small excerpt from a conversation
which is turned into a window through which the phenomenology (in a
non-technical sense) of a person’s social circumstances and experience is
captured and fleshed out in a compelling fashion, and in a manner which
resonates to the circumstances and experiences of many who might find
themselves in cognate circumstances.

Lecture 3 begins with another excerpt, but uses it largely as the point of
departure for a discussion of a type of sequence and of a characterizable locus
of interactional experience - the initial contact between someone calling on the
telephone and someone answering. The launching of the discussion from a
particular exemplar of an opening sequence imparts the flavor of empirical
analysis to the discussion, but in fact it is mainly near the end of the discussion
that Sacks takes up particulars of that initial fragment. In between his
characterization is chock full of the products of many empirical analyses, but
only their upshots are offered, with intendedly typicalized reports of conver-
sational exchanges to instantiate the themes, rather than analysis in each case,
for each observation or upshot, with specific instances or exemplars. Here
again Sacks catches the phenomenology of a social-interactional place in the
world, but whereas the place in the first two lectures was something like
“being a certain type of unique person in a setting,” here it is a transient
(though potentially recurrent) interaction state - answerer of the phone who

Introduction xlvii

may or may not be the “called,” and, if not, who may or may not get talked
to. 31

Lecture 4 is a specially striking exemplification of Sacks’ ability to
formulate an absolutely abstract issue, problem, or way of conceiving the
organization of talk, and then to use it to set a vernacularly characterizable
and recognizable class of occurrences into a relevant theoretical “space.” Here,
Sacks proposes to reconceive all utterances in (a) conversation in terms of three
possible “positions:” last, current, next, and he then begins a course of
theoretical observations about one of them - “next position” - as a purely
abstract possible object; and he finds, given how conversation seems empiri-
cally to be organized (especially given the turn-taking system which it seems
to employ) a set of characteristics of “next position” per se, characteristics
which will always have some particular embodiment by virtue of the
particular “current” utterance relative to which another is “next,” but which
are features of “next” position generically. From one such set of features -
that any “next” can accommodate some range of possible utterances or
utterance types, but not any utterance or utterance types - Sacks shows how
competition for a turn falls out as a consequence. For a possible next speaker
with something particular to say may see that it is possible to say it “next,”
but that each future “current” may restrict against this sayable in its next
position. Were things otherwise organized, a speaker with something to say
would never need to get a particular next position to say it in at the cost of
not getting to say it; everything “intended to be said” could, and perhaps
would, get said eventually — in some “any” next turn.

The power of this analytic tool is potentially very extensive, and some of it
made its way into the eventually published version of the turn-taking paper
(Sacks, Schegloff and Jefferson, 1974). More work along these lines was
planned; perhaps some day more will appear, however impoverished by
Sacks’ unavailability to press it ahead in his own distinctive way.

XII

After the Spring term of 1972 Sacks no longer recorded his lectures, and
made no special provision for circulating the work which he was teaching in

3 'i should remark that in this lecture - lecture 3 - more than any other place in the lectures,
there is a dialogue going on between Sacks and myself - my own part in it having been
developed first in my dissertation (1967) and the initial paper (Schegloff, 1968) drawn from
it, and then, most proximately to this lecture, in a revision of several chapters of the
dissertation for possible book publication, undertaken in the summer of 1970, and discussed
extensively with Sacks at the end of that summer. Some of that work has subsequently
appeared in modified form, e.g., in Schegloff, 1979 and 1986.

32 As noted earlier, at least some of Sacks’ lectures at the Linguistic Institute held during the
summer of 1973 were recorded, though Sacks did not choose to have them transcribed for
circulation. As well, Sacks continued to record many seminars and working sessions with
students and colleagues.

xlviii

Introduction

his classes. As it happened, I was that summer moving from a position at
Columbia University to one at UCLA, and for the next three years Sacks and
I maintained an often intensive, and intermittently attenuated, period of
collaborative work. Most of both Sacks’ sole-authored work and mine which
appeared over the following half dozen years was the delayed publication of
work done and written up much earlier. 33

Leaving aside for the moment work that was being newly launched or
developed in fresh directions during the years from Fall 1972 to Sacks’ death
in November, 1975, 34 those years saw the drafting of the paper on
turn-taking (Sacks, Schegloff and Jefferson, 1974), the earlier-mentioned
‘Two preferences . . .’ paper (Sacks and Schegloff, 1979), a paper on laughter
(Jefferson, Sacks and Schegloff, 1984) and the paper on repair (Schegloff,
Jefferson and Sacks, 1977). 35 An extensive account of the foci of work during
these years is beyond the scope of the present introduction. 36

Xlll

During the winter of 1974-5, Sacks and I were approached by several faculty
members at the University of California, Santa Barbara about the possibility
of establishing there an interdisciplinary program focussed on language,
discourse and interaction. We explored the possibility through the first half of
1975; we each visited the campus, gave talks, discussed the prospects with
the local interested faculty. It seemed increasingly clear that this was a serious
possibility, and that what was wanted was just the sort of enterprise that
conversation analysis was becoming - had already become. The prospects

33 Thus: Sacks, 1972a was the published version of Sacks' end product, dated June, 1965,
of what (rendered in more accessible language by David Sudnow) was Sacks’ Ph.D.
dissertation. Sacks, 1972b was a somewhat edited version of lectures from 19 66. Sacks,
1972c was originally a graduate student paper, written in 1962-3. Sacks, 1973 was the
published version of Sacks’ paper at the Georgetown Roundtable held in March, 1972. Sacks,
1974 was the published version of a paper delivered at a conference held in April, 1972.
Sacks, 1975 was the edited version of a lecture last given in 1968. Subsequent publications
under his name are edited versions of all or parts of pre-1972 course lectures, assembled by
Gail Jefferson (cf. introduction to Volume 1, p. ix, n. 1). Only Sacks, 1987 [1973], although
edited by others from a lecture, was first delivered after spring 1972.

Of co-authored papers, Schegloff and Sacks, 1973 was drafted (substantially in the form
in which it was published) in 1969.

34 Including his beginning to work with video materials, prompted in part during the 1973
Linguistic Institute by seeing the work of Charles and Marjorie Goodwin and its fit to
conversation analytic concerns.

35 Which Sacks and I outlined together in the spring of 1975, and which I then wrote the
intial draft of, after Sacks went off to the first Boston University Summer Session on
Ethnomethodology and Conversation Analysis.

36 After his death, I made a list of papers we had discussed more than once, and more than
casually, as needing to be drafted. There were 26 of them. Some account of these years may
yet be written.

Introduction

xlix

became increasingly enhanced. Jefferson (on the UCSB faculty that year)
could also be appointed; we could tailor a curriculum to the special character
of the subject matter and our approach to it; scheduling could be made
flexible; space for a collegial, “working group” arrangement was possible;
there would be support for our equipment needs, etc.

Finally, in the Fall of 1975, we received a request from those who were
guiding these developments at Santa Barbara. They wanted - quite reason-
ably - to know what we proposed to offer as a program in return for the
resources and possibilities which had been discussed over the preceding
months. Sacks and I had several informal conversations about this. Finally, in
mid-November, we decided we really had to sit down together and draw up
a serious plan to offer to Santa Barbara. We decided to meet at Irvine, in part
because Sacks had been suffering from an ear infection. We tentatively agreed
to meet on a Monday morning. When I called the Sunday night before the
scheduled meeting, the infection had not yet fully cleared up, and Sacks was
still taking medication for it. But he resisted the suggestion that we delay the
meeting. We would meet at the Irvine campus.

It was on his drive from the back canyons of Orange County to the Irvine
campus to discuss the specifics of the program in conversation analysis which
we might propose to Santa Barbara that his car was involved in a head-on
collision with a truck, and he was killed.

XIV

Reading the lectures now, and especially reading ones which entertain agendas
of work to be done (e.g., the last pages of Spring 1972, lecture 5), poses again
and again the question of where our understanding of language and talk, of
interaction and the social fates played out in it, of human sociality from the
most intimate emotion to the largest issues of social organization, where our
understanding would now be had Sacks not died in November, 1975.
Recalling the years immediately following the last of these lectures, when
some of that work was being advanced, and imagining what might have been
accomplished in a program designed to advance this undertaking, in a
supportive institutional environment, enhances the fantasy.

Whether or not the efforts of others succeed in establishing a discipline with
satisfactory payoffs and sustainable continuity, we shall not have the
discipline, or the understanding, which we would have had with him. Nor
will it avail for others literally to try to execute the plans of inquiry which he
projected. They were built from the breadth of his own past reading, from the
depth and range of his analytic and empirical work, and were the product of
the very special metier of his mind. What is needed is a continuous re-
energizing of inquiry by the example of his work and the possibilities which it
revealed - each person bringing to the enterprise the best mastery of past work
which they can achieve and the special contribution which the character of their
own talent makes possible. Not mechanical imitation or extrapolation but the

1

Introduction

best possible effort to advance the undertaking in original ways will constitute
the most appropriate and enduring celebration of Sacks’ contribution.

The first lecture presented in these volumes began with a consideration of
a conversation’s opening; the last ends with a puzzle about how much can be
infused into a conversation-opening “hello.” An astonishingly rich tapestry of
analysis comes between, in an intellectual career which did not tire of
repeatedly going back to the beginning, showing again and again that there
was an enterprise to be undertaken here. The achievement of the work is to
be found not only in its results, but in its prompting of an undertaking, and
in its constituting a standing invitation to others to join, and to begin, that
undertaking themselves.

Introduction

li

References

Albert, Ethel, 1964. Rhetoric, logic, and poetics in Burundi: cultural patterning of speech
behavior. American Anthropologist, 66, 6, part 2.

Atkinson, J. Maxwell and Drew, Paul, 1979. Order in Court (London: Macmillan).

Atkinson, J. Maxwell and Heritage, John C. (eds), 1984. Structures of Social Action
(Cambridge: Cambridge University Press).

Barker, Roger G. and Wright, HerbertF., 1951. One Boy's Day (New York: Harper and
Bros.).

Bilmes, Tack, 1988. The concept of preference in conversation analysis. Language in
Society, 17, 161-81.

Chafe, Wallace and Nichols, Johanna (eds), 1986. Evidentiality: the linguistic coding of
epistemology (Norwood, NJ: Ablex Publishing).

Davidson, Judy, 1984. Subsequent versions of invitations, offers, requests and proposals
dealing with potential or actual rejection. In JM Atkinson and JC Heritage (eds),
Structures of Social Action (Cambridge: Cambridge University Press).

Davidson, Judy, 1990. Modifications of invitations, offers and rejections. In George
Psathas (ed.), Interaction Competence (Washington, DC: University Press of America).

Garfinkel, Harold, 1967. Studies in Ethnomethodology ( Englewood Cliffs, NJ: Prentice
Hall).

Gellner, Ernest, 1975. Ethnomethodology: the re-enchantment industry or a Californian
way of subjectivity. Philosophy of the Social Sciences, 5, 4, 431-50.

Goffman, Erving, 1983. The interaction order. American Sociological Review, 48, 1-17.

Havelock, Eric A., 1963. Preface to Plato (Cambridge, Mass.: Harvard University Press).

Heritage, John C., 1984. Garfinkel and Ethnomethodology (Cambridge: Polity Press).

Homans, George C., 1961. Social Behavior: its elementary forms (New York: Harcourt,
Brace and World).

Jefferson, Gail, Sacks, Harvey, and Schegloff, Emanuel A., 1984. On laughter in pursuit of
intimacy. Working Papers, C(135), 1-20. Centro Intemazionale de Semistica e di
Linguistica, Urbino, Italy, Full version in Graham Button and John R. E. Lee (eds.) Talk
and Social Organization. 1987 (Clevedon, UK: Multi-liugual Matters) pp. 152-205.

Kroeber, Alfred, 1917. The superorganic. American Anthropologist, 19, 163-213.

Levinson, Stephen C., 1983. Pragmatics (Cambridge: Cambridge University Press).

Lord, Albert Bates, I960. The Singer of Tales (Cambridge, Mass: Harvard University
Press).

McLellan, David, 1977. Karl Marx: selected writings (Oxford: Oxford University Press).

Parry, Milman, 1971. The Making of Homeric Verse: the collected papers of Milman Parry,
edited by Adam Parry. (Oxford: The Clarendon Press).

Pomerantz, Anita, 1978. Compliment responses: notes on the co-operation of multiple
constraints. In Jim Schenkein (ed.). Studies in the Organization of Conversational
Interaction (New York: Academic Press), 79-112.

Pomerantz, Anita, 1984. Agreeing and disagreeing with assessments: some features of
preferred /dispreferred turn shapes. In J. M. Atkinson and J. C. Heritage (eds),
Structures of Social Action (Cambridge: Cambridge University Press), 57-101.

Sacks, Harvey, 1963. Sociological description. Berkeley Journal of Sociology, 8, 1-16.

Sacks, Harvey, 1972a. An initial investigation of the usability of conversational data for
doing sociology. In David N. Sudnow (ed.), Studies in Social Interaction (New York:
The Free Press), 31-74.

Sacks, Harvey, 1972b. On the analyzability of stories by children. In John J. Gumperz
and Dell Hymes (eds). Directions in Sociolinguistics (New York: Holt, Rinehart and
Winston), 325-45.

lii

Introduction

Sacks, Harvey, 1972c. Notes on police assessment of moral character. In David N.
Sudnow (ed.), Studies in Social Interaction (New York: The Free Press), 280-93.

Sacks, Harvey, 1973. On some puns with some intimations. In Roger W. Shuy (ed.).
Report of the 23rd Annual Roundtable Meeting on Linguistics and Language Studies
(Washington, DC: Georgetown University Press), 135-44.

Sacks, Harvey, 1974. An analysis of the course of a joke’s telling in conversation. In
Richard Bauman and Joel Sherzer (eds), Explorations in the Ethnography of Speaking
(Cambridge: Cambridge University Press), 337-53.

Sacks, Harvey, 1975. Everyone has to lie. In Ben Blount and Mary Sanchez (eds),
Sociocultural Dimensions of Language Use (New York: Academic Press), 57-80.

Sacks, Harvey, 1978. Some technical considerations of a dirty joke. In Jim Schenkein
(ed.), Studies in the Organization of Conversational Interaction (New York: Academic
Press), 249-70 [edited by Gail Jefferson from unpublished lectures].

Sacks, Harvey, 1987 [1973]. On the preferences for agreement and contiguity in
sequences in conversation. In Graham Button and John R. E. Lee (eds), 1987. Talk
and Social Organization (Clevedon: Multilingual Matters), 54-69.

Sacks, Harvey and Schegloff, Emanuel A., 1979. Two preferences in the organization of
reference to persons in conversation and their interaction. In George Psathas (ed.),
Everyday Language: studies in ethnomethodology (New York: Irvington), 15-21.

Sacks, Harvey, Schegloff, Emanuel A., and Jefferson, Gail, 1974. A simplest systematics
for the organization of turn-taking for conversation. Language, 50, 696-735. A
variant version also published in Jim Schenkein (ed.), Studies in the Organization of
Conversational Interaction (New York: Academic Press, 1978).

Schegloff, Emanuel A., 1967. The First Five Seconds: the order of conversational openings.
PhD dissertation, Department of Sociology, University of California, Berkeley.

Schegloff, Emanuel A., 1968. Sequencing in conversational openings. American Anthro-
pologist, 70, 1075-95.

Schegloff, Emanuel A., 1970. Openings sequencing and Answers. In The Social
Organization of Conversational Openings (MS).

Schegloff, Emanuel A., 1979. Identification and recognition in telephone conversation
openings. In George Psathas (ed.), Everyday Language: studies in ethnomethodology
(New York: Irvington).

Schegloff, Emanuel A., 1986. The routine as achievement. Human Studies, 9, 111-51.

Schegloff, Emanuel A., 1988. On an actual virtual servo-mechanism for guessing bad
news: a single case conjecture. Social Problems, 35, 442-57.

Schegloff, Emanuel A. and Sacks, Harvey, 1973. Opening up closings. Semiotica, 8,
289-327.

Schegloff, Emanuel A., Jefferson, Gail, and Sacks, Harvey, 1977 . The preference for
self-correction in the organization of repair in conversation. Language, 53, 361-82.

Schenkein, Jim (ed.), 1978. Studies in the Organization of Conversational Interaction
(New York: Academic Press).

Searle, John, 1986. Introductory essay: notes on conversation. In Donald G. Ellis and
William A Donahue (eds), Contemporary Issues in Language and Discourse Processes
(Hillsdale, NJ: Lawrence Erlbaum Associates).

Searle, John, 1992. Conversation reconsidered. In John R. Searle et al., (On) Searle on
Conversation (Amsterdam /Philadelphia: John Benjamin).

Taylor, Archer, 1931. The Proverb (Cambridge, Mass: Harvard University Press).

A Note on Some Effects of the
Taping, Transcribing, Editing, and
Publishing upon the Materials

Each process that the lectures have gone through, from taperecording to
typesetting, has had its effect on what appears in these volumes. As the
lectures were being taped, faulty equipment, excessive background noise, etc.,
could result in gaps within a recording, or the disappearance of an entire
lecture. This has provided an opportunity for the presentation of samples of
another instrument, as some of the absences have been filled in with materials
from Sacks’ research notes.

The transcribing process generated shoals of mis-hearings, only some of
which have been corrected. For example, a rendering of ‘Oral styles of
American folk narrators’ as “All styles . . . ”, or the transformation somehow
of ‘Von Neumann’ to “Baganin” were caught as the bibliographical
references were being assembled. A description, “the police car . . . follows
him along for a mile, finally pulls him over,” emerged on some nth reading
as much more likely to be “ . . . follows him along for awhile.” And a fresh
and knowledgable eye discovered that “. . . and they had a fantastic scene
with persons coming in ...” was almost certainly “. . . and they had a
fantastic team of persons coming in,” and rather more consequentially, that
“That’s in part the problem with Reichenbach’s second chapter” is in fact
reference to his “seventh” chapter, that a murky reference to data consisting
of “whatever it is that we have to have” was surely talking about “whatever
it is that we happen to have”, and reference to “the attending of a
prepositional phrase” was obviously reference to the “appending” of the
phrase. God only knows what further errors have slipped through.

There was never any requirement that the transcripts be verbatim, and
there is variation across and within them, although one small batch was
produced with a systematic concern for the very words. Not long after Sacks
was killed, a cache of taperecordings of his earliest lectures turned up. These
were treated, not as usual working tapes (to be transcribed as quickly as
possible and then tossed back into the pool of tapes for reuse), but as
something approaching a memorial. At attempt was made to capture as much
detail as possible; i.e., to transcribe them at the level of detail used on the
research materials, with Sacks’ frequent and prolonged silences, long drawn
out “uh”s, and very slight New York accent faithfully notated. But the
attempt was abandoned in the middle of the second page of transcript: At
that level of detail the lecture was simply not followable. It was necessary to
return to the standard format of the workaday lecture transcripts - the sense

Ixiii

lxiv

Note

of the specialness of this particular batch of tapes relegated to a commitment
to word for word accuracy. For the most part, however, some degree ol
spontaneous editing occurred at the transcription stage.

The editing per se ranges from faithful conservation of the text to
something very like wanton tampering. Now and again, remarks on
particular editorial effects can be found (see pp. ix-xi, 126-3 1, and 507-12),
but there is no comprehensive discussion.

As the volumes went to press, another series of changes occurred, geared to
bringing the materials more into line with standard literary usage, for
example, replacement of devices intended to emphasize the fact that these are
transcripts of spontaneous talk rather than written text; e.g., the rendering of
‘etc.’ as “etcetera” and (rather more variably) the rendering of numbers as
words, by the standard abbreviations and numerals. And, for example, the
various references to books and articles now look very much more like
standard bibliographical notes than spoken citations. A reference which went
into the typescript as “I come by this sort of consideration via a paper written
by a fellow named Richard Gunter, 19 66, Journal of Linguistics called ‘On
accents in dialogue,’” became “I come by this sort of consideration via a
paper by Richard Gunter, ‘On accents in dialogue,’ Journal of Linguistics
(1966).” A typescript entry, reference to a review article “in a book called
Studies in Language and Literature, edited by A. Marquart, 1954, called
‘English sentence connectors,’ by Seymour Chapman. That’s on page 315,”
came back as reference to a review article “by Seymour Chapman, ‘English
sentence connectors,’ in A. Marquart (ed.), Studies in Language and Literature
(1954), p. 315.” As a sort of compromise between good form and actual
occurrence, the non-initial reference to the author, in this and other citations
throughout the volumes, has been restored.

Acknowledgements

With thanks to David Sudnow who kick-started the editing process when it
had stalled, and to Betty Sacks who kept it going. And to lan Anderson, J.
Maxwell Atkinson, Tom Boda, Judy Davidson, Paul Drew, Robert Dunstan,
Konrad Ehlich, B.J. Fehr, Rich Frankel, Jo Ann Goldberg, Charles Goodwin,
Marjorie Goodwin, Auli Hakulinen, Paul ten Have, Christian Heath, John
Heritage, Hanneke Houtkoop, Willem Houtkoop, Eva Konig, Jerry
Krakowski, John R.E. Lee, Diane Lee, Gene Lerner, Jennifer Mandelbaum,
Michael Moerman, Bruce Peddy, Anita Pomerantz, George Psathas, Blaine
Roberts, Joan Sacks, Jim Schenkein, Albert Stuulen, Alene Terasaki, Roy
Turner, Rod Watson, and those known only as ‘transcriber unknown’.

The editor and publishers wish to thank the following for permission to use
copyright material: Macmillian Publishing Company for an extract from The
Urban Villagers by Herbert J. Gans; Mouton de Gruyter for an extract from
Language in the Crib by Ruth Weir; Oxford University Press for an extract
from ‘Is everyday language inconsistent?’ by Avram Stroll, in Mind, LXIII,
no. 250. Every effort has been made to trace copyright holders of extracted
material and the publishers would be grateful to hear from any individuals or
institutions whose clearance has not been obtained at the date of publication.

lxv

Part I

Fall 1964-Spring 1965

Lectures on Conversation, Volume I, II Harvey Sacks
© 1995 The Estate of Harvey Sacks. ISBN: 978-1-557-86705-6

Lecture 1

Rules of Conversational Sequence

I’ll start off by giving some quotations.

(1) A: Hello
B: Hello

(2) A: This is Mr Smith may I help you
B: Yes, this is Mr Brown

(3) A: This is Mr Smith may I help you
B: I can’t hear you.

A: This is Mr Smith .

B: Smith.

These are some first exchanges in telephone conversations collected at an
emergency psychiatric hospital. They are occurring between persons who
haven’t talked to each other before. One of them, A, is a staff member of this
psychiatric hospital. B can be either somebody calling about themselves, that
is to say in trouble in one way or another, or somebody calling about
somebody else.

I have a large collection of these conversations, and I got started looking at
these first exchanges as follows. A series of persons who called this place
would not give their names. The hospital’s concern was, can anything be done
about it? One question I wanted to address was, where in the course of the
conversation could you tell that somebody would not give their name? So I
began to look at the materials. It was in fact on the basis of that question that
I began to try to deal in detail with conversations.

I found something that struck me as fairly interesting quite early. And that
was that if the staff member used “This is Mr Smith may I help you” as their
opening line, then overwhelmingly, any answer other than “Yes, this is Mr
Brown” (for example, “I can’t hear you,” “I don’t know,” “How do you

A combination of Fall 1964, tape 1, side 2 and tape 2, side 1, with brief extracts
from Winter 1965, lecture (1) - the parenthesis indicate that the original transcripts
were unnumbered, the current numbering likely but not certain - pp. 1 and 11-12
(transcriber unknown) and Spring 1965 (’64-’65), lecture 3, pp. 6-7 (transcriber
unknown).

The lectures’ titles are intended to give a handle on them, and only partially
capture the contents.

Lectures on Conversation, Volume I, II Harvey Sacks
© 1995 The Estate of Harvey Sacks. ISBN: 978-1-557-86705-6

3

4

Part I

spell your name?”) meant that you would have serious trouble getting the
caller’s name, if you got the name at all.

I’m going to show some of the ways that I’ve been developing of analyzing
stuff like this. There will be series of ways fitted to each other, as though one
were constructing a multi-dimensional jigsaw puzzle. One or another piece
can be isolated and studied, and also the various pieces can be studied as to
how they fit together. I’ll be focussing on a variety of things, starting off with
what I’ll call ‘rules of conversational sequence.’

Looking at the first exchange compared to the second, we can be struck by
two things. First of all, there seems to be a fit between what the first person
who speaks uses as their greeting, and what the person who is given that
greeting returns. So that if A says “Hello,” then B tends to say “Hello.” If
A says“This is Mr Smith may I help you,” B tends to say “Yes, this is Mr
Brown.” We can say there’s a procedural rule there, that a person who speaks
first in a telephone conversation can choose their form of address, and in
choosing their form of address they can thereby choose the form of address the
other uses.

By ‘form’ I mean in part that the exchanges occur as ‘units.’ That is,
“Hello” “Hello” is a unit, and “This is Mr Smith may I help you” “Yes, this
is Mr Brown” is a unit. They come in pairs. Saying “This is Mr Smith may
I help you” thereby provides a ‘slot’ to the other wherein they properly would
answer “Yes, this is Mr Brown.” The procedural rule would describe the
occurrences in the first two exchanges. It won’t describe the third exchange,
but we’ll come to see what is involved in such materials.

Secondly, if it is so that there is a rule that the person who goes first can
choose their form of address and thereby choose the other’s, then for the unit,
“This is Mr Smith may I help you” “Yes, this is Mr Brown,” if a person uses
“This is Mr Smith ...” they have a way of asking for the other’s name -
without, however, asking the question, “What is your name?” And there is
a difference between saying “This is Mr Smith may I help you” - thereby
providing a slot to the other wherein they properly would answer “Yes, this
is Mr Brown” - and asking the question “What is your name?” at some point
in the conversation. They are very different phenomena.

For one, in almost all of the cases where the person doesn’t give their name
originally, then at some point in the conversation they’re asked for their name.
One way of asking is just the question “Would you give me your name?” To
that, there are alternative returns, including “No” and “Why?” If a caller
says “Why?” the staff member may say something like, “I want to have
something to call you” or “It’s just for our records.” If a caller says “No,”
then the staff member says “Why?” and may get something like “I’m not
ready to do that” or “I’m ashamed.”

Now, I’ll consider many times the use of “Why?” What I want to say
about it just to begin with, is that what one does with “Why?” is to propose
about some action that it is an ‘accountable action.’ That is to say, “Why?”
is a way of asking for an account. Accounts are most extraordinary. And the
use of accounts and the use of requests for accounts are very highly regulated

Lecture 1

5

phenomena. We can begin to cut into these regularities by looking at what
happens when “May I have your name?” is followed by “Why?” Then you
get an account; for example, “I need something to call you.” The other might
then say, “I don’t mind.” Or you might get an account, “It’s just for our
records.” To which the other might say, “Well I’m not sure I want to do
anything with you, I just want to find out what you do” - so that the records
are not relevant.

What we can see is that there are ways that accounts seem to be dealable
with. If a person offers an account, which they take it provides for the action
in question being done - for example, the caller’s name being given - then if
the other can show that the interest of that account can be satisfied without
the name being given, the name doesn’t have to be given. That is, if the
account is to control the action, then if you can find a way that the account
controls the alternative action than it proposed to control, you can use it that
way.

It seems to be quite important, then, who it is that offers the account.
Because the task of the person who is offered the account can then be to, in
some way, counter it. Where, alternatively, persons who offer an account seem
to feel that they’re somehow committed to it, and if it turns out to be, for
example, inadequate, then they have to stand by it.

The fact that you could use questions — like “Why?” — to generate
accounts, and then use accounts to control activities, can be marked down as,
I think, one of the greatest discoveries in Western civilization. It may well be
that that is what Socrates discovered. With his dialectic he found a set of
procedures by which this thing, which was not used systematically, could
become a systematic device. Socrates will constantly ask “Why?,” there will
be an answer, and he’ll go on to show that that can’t be the answer. And that
persons were terribly pained to go through this whole business is clear enough
from the Dialogues. And it’s also clear in our own experiences. And in the
materials I’ll present.

We see, then, one clear difference between providing a slot for a name, and
asking for a name. Asking for a name tends to generate accounts and
counters. By providing a slot for a name, those activities do not arise.

We can also notice that, as a way of asking for the other’s name, “This is
Mr Smith . . .” is, in the first place, not an accountable action. By that I mean
to say, it’s not required that staff members use it and they don’t always use
it, but when they do, the caller doesn’t ask why. “This is Mr Smith . . .” gets
its character as a non-accountable action simply by virtue of the fact that this
is a place where, routinely, two persons speak who haven’t met. In such places
the person who speaks first can use that object. And we could say about that
kind of item that the matters discriminated by its proper use are very
restricted. That is to say, a call is made; the only issue is that two persons are
speaking who presumably haven’t met, and this object can be used.

Furthermore, the matters are discriminated in different terms than those
which the agency is constructed for. That is, they are discriminated in terms
of ‘two people who haven’t met’ rather than, for example, that an agency staff

6

Part 1

member is speaking to someone calling the agency for help. And where one
has some organization of activities which sets out to do some task - and in this
case it’s important for the agency to get names - then if you find a device
which discriminates in such a restricted fashion, you can use that device to do
tasks for you.

Now, given the fact that such a greeting as “This is Mr Smith
provides for the other giving his own name as an answer, one can see what the
advantage of “Hello’’ is for someone who doesn’t want to give their name.
And I found in the first instance that while sometimes the staff members use
“Hello’’ as their opening line, if it ever occurred that the persons calling the
agency spoke first, they always said “Hello.”

Persons calling could come to speak first because at this agency, caller and
staff member are connected by an operator. The operator says “Go ahead
please” and now the two parties are on an open line, and one can start talking
or the other can start talking. This stands in contrast to, for example, calling
someone’s home. There, the rights are clearly assigned; the person who
answers the phone speaks first. If they speak first, they have the right to
choose their form. If they have the right to choose their form, they have the
right thereby to choose the other’s. Here, where the rights are not clearly
assigned, the caller could move to speak first and thereby choose the form.
And when callers to this agency speak first, the form they choose is the unit
“Hello” “Hello.” Since such a unit involves no exchange of names, they can
speak without giving their name and be going about things in a perfectly
appropriate way.

Now, there are variant returns to “This is Mr Smith may I help you?” one
of which is in our set of three exchanges: “I can’t hear you.” I want to talk
of that as an ‘occasionally usable’ device. That is to say, there doesn’t have to
be a particular sort of thing preceding it; it can come at any place in a
conversation. Here is one from the middle of a conversation, from a different
bunch of materials.

A: Hey you got a cigarette Axum. I ain’t got, I ain’t got a good cigarette,

and I can’t roll one right now. Think you can afford it maybe?

B: I am not here to support your habits.

A: Huh? My helplessness?

B: I am not responsible for supporting your habits ( )

A: My habits ((laughing))

Our third exchange from the psychiatric hospital has the device used at the
beginning of the conversation.

A: This is Mr Smith may I help you

B: I can’t hear you.

A: This is Mr Smith .

B: Smith.

Lecture 1

7

What kind of a device is it? What you can see is this. When you say “I
can’t hear you,” you provide that the other person can repeat what they said.
Now what does that repetition do for you? Imagine you’re in a game. One of
the questions relevant to the game would be, is there a way in that game of
skipping a move? It seems that something like ”1 can’t hear you” can do such
a job. If you introduce it you provide for the other to do some version of a
repeat following which you yourself can repeat. And then it’s the other’s turn
to talk again. What we find is that the slot where the return would go - your
name in return to “This is Mr Smith ...” - never occurs.

It is not simply that the caller ignores what they properly ought to do, but
something rather more exquisite. That is, they have ways of providing that
the place where the return name fits is never opened. So that their name is not
absent. Their name would be absent if they just went ahead and talked. But
that very rarely occurs. The rules of etiquette - if you want to call them that,
though we take etiquette to be something very light and uninteresting and to
be breached as you please - seem to be quite strong. Persons will use ways to
not ignore what they properly ought to do by providing that the place for
them to do it is never opened.

I hope it can also be seen that a device like “I can’t hear you” - the repeat
device, providing for a repetition of the thing that was first said, which is then
repeated by the person who said “I can’t hear you” - is not necessarily
designed for skipping a move. It is not specific to providing a way of keeping
in the conversation and behaving properly while not giving one’s name. It can
be used for other purposes and do other tasks, and it can be used with other
items. That’s why I talk about it as an ‘occasional device.’ But where that is
what one is trying to do, it’s a rather neat device.

Let me turn now to a consideration which deals with a variant return to
“May I help you?” That is, not “Yes . . .” but “I don’t know.” I’ll show a
rather elaborate exchange in which the staff member opens with a version of
“This is Mr Smith may I help you” but the combination gets split. The name
is dealt with, and when the “can I help you” is offered, it occurs in such a way
that it can be answered independent of the name. 1

Op : Go ahead please

A: This is Mr Smith (13: Hello) of the Emergency Psychiatric Center can

I help you.

B: Hello?

A: Hello

B: I can’t hear you.

A: I see. Can you hear me now?

B: Barely. Where are you, in the womb?

1 The fragment of data is reproduced pretty much as Sacks transcribed it,
preserving his attempts to deal with simultaneous talk (i.e., A: This is Mr Smith ( B :
Hello) of the Emergency Psychiatric Center) and silence (e.g., B: I uh Now that
you’re here . . .). See lecture 9, pp. 66 and 68 for two other approaches by him to
simultaneous talk in this same conversation.

8

Part I

A: Where are you calling from?

B: Hollywood.

A: Hollywood.

B: I can hear you a little better.

A: Okay. Uh I was saying my name is Smith and I’m with the

Emergency Psychiatric Center.

B : Your name is what?

A: Smith.

B : Smith?

A: Yes.

A: Can I help you?

B\ I don’t know hhheh I hope you can.

A: Uh hah Tell me about your problems.

B: I uh Now that you’re here I’m embarassed to talk about it. I don’t

want you telling me I’m emotionally immature ’cause I know I am

I was very puzzled by “I don’t know” in return to “May I help you.” I
couldn’t figure out what they were doing with it. And the reason I was
puzzled was that having listened to so many of these things and having been
through the scene so many times, I heard “May I help you” as something like
an idiom. I’m going to call these idiom-like things ‘composites.’ That means
you hear the whole thing as a form, a single unit. And as a single unit, it has
a proper return. As a composite, “May I help you” is a piece of etiquette; a
way of introducing oneself as someone who is in the business of helping
somebody, the answer to which is “Yes” and then some statement of what it
is one wants. We can consider this item in terms of what I’ll call the ‘base
environment’ of its use.

By ‘base environment’ I mean, if you go into a department store,
somebody is liable to come up to you and say “May I help you.” And in
business-type phone calls this item is routinely used. And if you come into a
place and you don’t know what it’s like, and somebody comes up to you and
uses such an item, that’s one way of informing you what kind of a place it is.
So, if a new institution is being set up, then there are available in the society
whole sets of ways that persons go about beginning conversations, and one
could, for example, adopt one or another of a series of them as the ones that
are going to be used in this place.

Now the thing about at least some composites is that they can be heard not
only as composites, but as ordinary sentences, which we could call ‘construc-
tives,’ which are understood by taking the pieces and adding them up in some
way. As a composite, “May I help you” is a piece of etiquette, a signal for
stating your request - what you want to be helped with. Alternatively, as a
constructive, “May I help you” is a question. If one hears it as a question, the
piece of etiquette and its work hasn’t come up, and “I don’t know” is a
perfectly proper answer.

Further, “I don’t know” may be locating a problem which “May I help
you” is designed, in the first place, to avoid. In its base environment, for

Lecture 1

9

example a department store, it’s pretty much the case that for a customer, the
question of whether some person “can help’’ is a matter of the department
store having made them the person who does that. That is to say, lots of
things, like telling you whether you can find lingerie in a certain size, is
something anybody can do, and as long as the department store says this
person is going to do it, that’s enough. But we’re dealing with a psychiatric
hospital. In a department store, being selected to do a job and having
credentials to do it are essentially the same thing. In a psychiatric hospital and
lots of other places, however, they are very different things. That is, whether
somebody can help you if you have a mental disorder, is not solved or is not
even presumptively solved by the fact that they’ve been selected by somebody
to do that job. The way it’s solved in this society is by reference to such things
as having been trained in a particular fashion, having gotten degrees, having
passed Board examinations, etc.

Now, in the base environment of the use of “May I help you?’’ there is, as
I say, no difference essentially between having credentials and being selected.
If one can formulate the matter in a psychiatric hospital such that those things
come on as being the same, then one needn’t start off by producing one’s
credentials at the beginning of the conversation. And in my materials, again
and again, when “May I help you” is used the person calling says “Yes” and
begins to state their troubles.

As a general matter, then, one can begin to look for kinds of objects that
have a base environment, that, when they get used in that environment
perform a rather simple task, but that can be used in quite different
environments to do quite other tasks. So, a matter like ‘credentials’ can be
handled by this “May I help you” device. There will be lots of other devices
which have a base environment, which do some other task in some other
environment.

Before moving off of “May I help you” I want to mention one other thing
about it. If the base environment is something like a department store, then,
when it’s used in other places - for example, a psychiatric hospital - one of
the pieces of information it seems to convey is that whatever it is you propose
to do, you do routinely. To whomsoever that calls. That is, it’s heard as a
standardized utterance. How is that relevant? It can be relevant in alternative
ways. First of all, it can be a very reassuring thing to hear. Some persons feel
that they have troubles, and they don’t know if anybody else has those
troubles; or, if others do have those troubles, whether anybody knows about
them. If someone knows about them, then there may be a known solution to
them. Also and relatedly, a lot of troubles - like mental diseases - are things
that persons feel very ambivalent about. That is, they’re not sure whether it’s
some defect of their character, or something else. That, in part, is why they’re
hesitant to talk about it. And it seems that one of the ways one begins to tell
people that they can talk, that you know what they have and that you
routinely deal with such matters, is to use manifestly organizational talk.

“May I help you,” then, can be a reassuring way to begin. It can
alternatively be something else. Consider the exchange I just showed, in which

10

Part I

such standardized utterances as “May I help you” and “Tell me about your
problems” are used.

A: Can I help you?

B: I don’t know hhheh I hope you can

A: Uh hah Tell me about your problems

B: I uh Now that you’re here I’m embarrassed to talk about it. I

don’t want you telling me I’m emotionally immature ‘cause I know
I am

That is, the use of standardized, manifestly organizational talk can provide for
the person calling that they’re going to get routine treatment. But ‘routine’,
for them, may not be such a happy thing. Because, for example, they’ve been
through it before. But they may have gone through it, as psychiatrists would
say, part way. For example, they were in analysis for three years and ran out
of money, or the psychiatrist wouldn’t keep them on, or they didn’t want to
stay. Part way, they may have come to some point in the analysis where they
‘knew what was wrong with them.’ That is, they knew the diagnostic term.
But that diagnostic term may have had a lay affiliate. By that I mean, if a
psychiatrist says you’re regressed, it’s a technical term. But ‘regressed’ is also
a lay term, and as a lay term it doesn’t have a great deal of attractiveness. If
one finds oneself living with a lay understanding of such a term, where the
term is not a very nice thing to have in its lay sense, then when you hear
someone using such an item as “May I help you,” you can hear that some
procedure will be gone through, the upshot of which will be the discovery
of what you ‘already know’ - the knowing of which doesn’t do you any
good.

Related to that are such things as, some people seem to feel very much
disturbed about the fact that their relationship to a psychiatrist or to other
doctors is monetary. What they want, they say, is a personal solution. Ask
them what they want, “Well, that you don’t have to pay for it.” When they
hear “May I help you,” they hear ‘a professional.’ But they feel that the way
you get cured is by getting an affiliation to somebody which is like the
affiliations that they failed to get in their lives. That is, they may already have
come to learn from some other psychiatrist that the failure of love by their
parents is the cause of their troubles. Then, what they come to see is that they
need the love of somebody else. And they can’t get that from a therapist.
Because as soon as they don’t pay, that’s the end of the relationship.

Now let me just make a few general points. Clearly enough, things like
“This is Mr Smith,” “May I help you?” and “I can’t hear you” are social
objects. And if you begin to look at what they do, you can see that they, and
things like them, provide the makings of activities. You assemble activities by
using these things. And now when you, or I, or sociologists, watching people
do things, engage in trying to find out what they do and how they do it, one
fix which can be used is: Of the enormous range of activities that people do,
all of them are done with something. Someone says “This is Mr Smith” and

Lecture 1

11

the other supplies his own name. Someone says “May I help you’’ and the
other states his business. Someone says “Huh?” or “What did you say?” or
“I can’t hear you,” and then the thing said before gets repeated. What we
want then to find out is, can we first of all construct the objects that get used
to make up ranges of activities, and then see how it is those objects do get
used.

Some of these objects can be used for whole ranges of activities, where for
different ones a variety of the properties of those objects will get employed.
And we begin to see alternative properties of those objects. That’s one way we
can go about beginning to collect the alternative methods that persons use in
going about doing whatever they have to do. And we can see that these
methods will be reproducible descriptions in the sense that any scientific
description might be, such that the natural occurrences that we’re describing
can yield abstract or general phenomena which need not rely on statistical
observability for their abstractness or generality.

There was a very classical argument that it would not be that way; that
singular events were singular events, given a historian’s sort of argument, that
they just happen and they get more or less accidentally thrown together. But
if we could find that there are analytically hard ways of describing these things
— where, that is, we re talking about objects that can be found elsewhere, that
get placed, that have ways of being used; that are abstract objects which get
used on singular occasions and describe singular courses of activity - then
that’s something which is exceedingly non-trivial to know.

One final note. When people start to analyze social phenomena, if it looks
like things occur with the sort of immediacy we find in some of these
exchanges, then, if you have to make an elaborate analysis of it - that is to say,
show that they did something as involved as some of the things I have
proposed - then you figure that they couldn’t have thought that fast. I want
to suggest that you have to forget that completely. Don’t worry about how
fast they’re thinking. First of all, don’t worry about whether they’re
‘thinking.’ Just try to come to terms with how it is that the thing comes off.
Because you’ll find that they can do these things. Just take any other area of
natural science and see, for example, how fast molecules do things. And they
don’t have very good brains. So just let the materials fall as they may. Look
to see how it is that persons go about producing what they do produce.

Lecture 2

On Suicide Threats Getting
Laughed Off

Here are some lines that occurred in one of the conversations I collected. This
is a woman talking.

A: But about two months ago I was still home on uh one Sunday, oh we

had five children and I got home from church and he’s got a butcher
knife. He told the kids to go to the park and play. This is kind of
unusual for him because he doesn’t like them, especially the baby, to
go anywhere unless we’re there.

B- Aha.

A: After they were all gone, I was laying on the couch just reading the

Sunday paper and he came over there and started holding this butcher
knife at my throat. And I said what is the matter with you. He said
I’m going to kill you. I’m going to end it all. And I said oh for
goodness sake put it down and go.

— H started to laugh it ofF. And he sat there for about an hour. So I
thought well, he kept threatening to kill me. And then he would pull
it back as if to stab me. And I just laid there and prayed. I almost
believed he was crazy.

And then he had been acting fairly good since then. He doesn’t have
any religion and I’m Catholic. But I said why don’t you go down and
talk to the priest. Maybe he would help you.

Here’s another, from the same conversation.

A: What if you won’t come. I mean how do I- about- Oh, the last time

he tried to kill me he sat and wrote a long suicide note or whatever.
I don’t know. I didn’t read it. This was on a Sunday when the kids
and I got home from church and he wanted to know if I went to
church with the kids and they said of course. She always goes to
church with us. He said I know she’s got a boyfriend. I said quit
acting silly in front of the kids. What’s the matter with you. He says
oh, and then, I don’t know. Anyway, this time he tried to kill me. He
wrote this long note.

—+l just acted like I thought he was kidding. I didn’t want him to think

Fall 1964, tape 5, side 1

12

Lecture 2

13

I was taking him seriously. He said well Joey run down to the police
station before I do something I don’t want to do. I said Daddy quit
it. Joey says Daddy I don’t want to go down there, they’ll all look
funny at me when they read the note. I says Joey run outside, Daddy ’s
only kidding. He says no I’m not. You’d better let me do it. Then he
got in the car and went tearing off. I looked for the note last night and
he didn’t have one, so I thought oh maybe he knew I’d wake up and
maybe not. But I don’t want to leave it go.

Here’s another, from an altogether different conversation.

A: I mean the thing that makes it even more serious to me is the once or

twice that I’ve mentioned it, not deliberately, but kind of slipping, to
the family or anything like that, they try to make a joke of it, you
know,

B: Well no, see, here we take all of that seriously.

A: And believe me it’s no joke because as I say I just don’t feel my life

is worth anything at this point.

B: Well we take that very seriously and when someone feels that way we

try to do whatever we can to try to help them work out of that feeling.
And we’d like to help you.

A: Okay, fine then.

And another conversation.

A: I want somebody to talk me out of it, I really do.

B: Uh huh,

A: But I can’t call any of my friends or anybody, ’cause they’re just going

to say oh that’s silly or that’s stupid.

B: Uh huh,

A: I guess what you really want is someone to say, yes I understand why

you want to commit suicide, I do believe you.

Recurrently in these conversations, persons say that when they use the line
“I’m going to kill myself,” others laugh. And that’s not only by self-report,
I have things from police reports of suicides where the police then ask persons
around, “Well, did they ever threaten to kill themselves?” and those persons
say “Well, he said he was going to kill himself but we just laughed it off.”
And the question I began to address was, what kind of relationship was there
between the statement “I’m going to kill myself,” and laughter. How is it
that laughter would be done there?

Okay, let’s hold that problem now and turn to another set of materials, via
which we’ll be trying to see what might be involved in it. I said about the
opening lines of conversations that they seem to come in pairs. And that one
person could choose the form of greeting he used. And that if one person

14

Part I

could choose their own they could choose the other’s. Now it seems that there
is a general class of such kinds of things, which I’m going to call ‘ceremonials.’
Other examples are, for example, “How are you feeling?” to which you
return “Fine.” If one person, then, uses a ceremonial, the other properly
returns with a ceremonial.

Let’s look at “How are you feeling?” It’s routinely used between persons
as either a greeting or greeting substitute. And it’s used between persons who
needn’t have very much intimacy. But there is a smaller group of persons
included in the circle of persons who routinely use this object. Call the larger
group ‘others’ and the smaller, a special class of others. I won’t at this point
go into describing in detail what the properties are of this special class of
others. Roughly, they are persons who, if one has a trouble, one turns to them
for help. Without giving some of the ways we could talk about their relation
to some ‘one’ - call that one Ego - like, for example, they may be kin, I want
to approach it in a little different way. One of the ways they stand to each
other is, if something happens to Ego, then, whoever it is that might be trying
to discover why that thing happened, could refer for explanations to these
others. So let’s say they’re ‘causally bound’ to the person who may have
trouble.

And that could quite easily make it apparent how it is that if such a one
is turned to for help, they have a feeling of obligation. They would have a
feeling of obligation by virtue of the fact that if, let’s say, a suicide occurs,
then, even if they hadn’t been approached for help in the first place, the
question would be asked, well what was up with that family that she should
have killed herself? Many things that might happen to Ego will be causally
explained by virtue of something that the other did. And if others want to
avoid that happening to them, then when some Ego turns to them, they feel
like giving help. And of course the fact that these others walk around with all
kinds of guilt turns in part on that causal relationship. Now this is among
laymen; you don’t have to have scientific theories to feel this causal
involvement. Any layman would ask, if somebody says “My brother killed
himself,” “Well what’s the matter with the family?” That’s where you would
look for the source.

Further, somebody who is not a part of this small group of others can
become causally involved by virtue of the fact that Ego has asked them for
help in some way and been turned down. If something then happens to Ego,
it seems that even if you aren’t one of that small group of others, you know
about the fact that Ego was troubled, how come you didn’t do anything? So
knowledge of the trouble is often sufficient to bring one into causal
involvement.

Now these people, the whole circle, are going around constantly saying
“How are you feeling?” Properly, the return is “Fine.” And this can be fairly
dramatic. I’ve sat around in hospitals, and in a hospital persons who are, say,
recuperating from serious diseases may be sitting in wheelchairs outside their
room or in the common room, etc. A doctor walks by a person who looks like
they’re just about to go, and says “How are you feeling?” and they say

Lecture 2

15

“Fine.” Sometimes, however, a person may take that “How are you feeling?”
and attempt to use it to present their troubles. And one sort of thing that
happens in that case is that persons who listen when somebody begins to tell
them their troubles, talk about themselves routinely as ‘softhearted,’ ‘fools,’
and that sort of thing. And when persons talk about themselves as softhearted
with respect to others, it’s probably something like this that’s happened to
them: They listen, then they find themselves ‘involved.’ Involved, however,
without the basic properties that would initiate their relevant obligation, but
not knowing what to do. And not knowing how to get out, either, because
they ‘know too much.’

On the other hand, the fact that there is that ceremonial relation between
“How are you feeling?” and “Fine,” may set up the following situation.
Routinely, if you look at first interviews (and perhaps later interviews also)
between psychiatrists and patients or possible patients, they begin like this:

A: How are you feeling?

B: It’s a long story.

A: That’s alright, I have time.

What is this “It’s a long story,” and things like it, doing here? The person
knows that the line “How are you feeling?” is a ceremonial line, and it’s a
breach of the proper forms to begin to launch right then and there into what
it is that’s bothering you. So what they then do is try to initiate another
ceremony which would provide the basis for them talking. Typically this
other ceremony is nicely done, in that what one does is offer a tentative refusal,
like “It’s a long story” or “It’ll take hours,” which turns it back to the other,
referring to their circumstances; for example, their schedule. And it invites the
other to then say that their schedule does not control your activities, so go
ahead and talk.

Now, persons who are causally bound are obliged to give help when help
is asked for. That means in part that they’re in bad shape if they don’t give
help and trouble occurs. They’re responsible for someone. Others hold them
responsible, and they feel responsible. The question is, is there some way that
they can go about refusing to give help without ‘refusing,’ in the same way
that I’ve talked about refusing to give one’s name without ‘refusing’? One
solution would be to find a way to set up the first remark as the first remark
of a ceremonial. Because then the proper return is a ceremonial. While there
are some ceremonials that come off strictly by virtue of the particular object
that’s used, there are others that are classes of ceremonials. Three common
classes are jokes, games, performances. They all have the character that the
next move - or some other given move in the sequence — is the end of it, and
that’s the end of the whole thing. You tell a joke, there’s a laugh. A game has
a set beginning and end. A play has the same character.

That is, I think it’s the fact that we have ceremonial relationships between
various objects and their proper returns, that sets up the sort of business we
started off with: “I’m going to kill myself’ followed by laughter. When

16

Part 1

somebody says “I’m going to kill myself,” if the other can cast it into one of
the ceremonial forms, then that can end the interchange. One wouldn’t
have heard the ‘cry for help.’ One would have heard a joke. And one would
have behaved properly with respect to a joke. And it appears that,
alternatively to giving help, one gets cases of just those three common classes
of ceremonials. Somebody laughs, or they say “Nice performance,” or “Quit
playing.” And that would provide, then, for closing that thing off without,
however, having been in the situation of refusing help in the sense of saying
“no,” or other such things. So we can see how that form provides for this
thing to happen.

We can also see how awfully painful it must be for persons who are deeply
troubled, and who constantly have people coming up to them and saying
“How are you feeling?” when they can’t come back. Now and then we see
that very problem referred to in a joking form. Here is an instance.

A: How are you feeling?

B: You really want to know? ha ha

A: ha ha

That is, someone, asked “How are you feeling?” jokingly proposes: What if
I were to take this, not as a ceremonial form, but as a serious invitation. Then
where would you be? And when people are asked “Well why don’t you tell
somebody?” they say “It’d be like a melodrama!” or “How can you tell them,
they’ll just laugh!”

I want to say another thing about ceremonials. Here is something very nice.
Very lovely. Lovely in a way, but quite awful, also. When I was thinking
about this stuff, I came across a very frequently recurring kind of statement.
I’ll just give one case; a long extract in which a widow is telling a psychiatrist
of some problems she is having with her married daughter.

A: Well, I’ll tell you really what got into me last week. You know I was

just talking about Thanksgiving beingThursday, and she had to
— ^prepare, but she didn’t invite me. And I go home and I start to think
about it, and you know, when I spoke to you alone there a couple of
minutes, I shouldn’t have talked about that, because there was
something else that was- I mean I touched on it, but there was
something else.

- *T just had a feeling that I wasn’t wanted anymore in their house. At
least by her husband. Naturally she can’t do anything about it. You
know, I mean if she could, she would start fighting with him, and I
wouldn’t want to be a cause of that you know. But I thought that
because, when I first went to the doctor that I went to, this internist
I was going to last July, and she suggested that I go to a doctor in the
Valley that she knew. She says well, it’s a good idea because if you
have to be hospitalized, or if you’re depressed or anything, you could
stay with me for a couple of days. She says I can’t get down to see you

Lecture 2

17

that often, with the children. But I’ll take care of you if you stay at my
house.

So this is in July. And I wasn’t able to go to him because I didn’t have
the money to go. I finally in October, had to go to the hospital. And
I was there for three days and got these tests, which just made me
awfully weak, and when I got out she called for me because she had
my car while I was in the hospital. She called for me and didn’t
—••ask if I wanted to stay over that night. I get out of the hospital and
I have no- and I have to drive home, and I felt so weak by myself.
I mean, she couldn’t because otherwise she couldn’t have gotten back.
I mean it was just one of those things. But the better thing would have
been if I could have stayed up there at least overnight and when I felt
—’•fresh, take the trip down. But she didn’t even ask. And I know it isn’t
like Lila not to ask, when this was the original reason for my going up
there. And I just know that she was warned that she better not bring
me home.

And of course I started feeling sorry for myself. And then, when we
were there Wednesday, she said something about preparing a
—’•Thanksgiving dinner next Thursday and she didn’t say anything
about me. I figured, well, instead of the family, which we always had,
the family together, it’s not at my house, it’s at her house. I mean
during the time I was married I used to have seventeen or twenty for
dinner because the whole family. And then she had taken over lately.
So I thought well, maybe she’s gonna have her son. And it’s not up
to me to expect her to have me. And then I thought well, maybe she
figured Jay

that’s the son

is going to be there, and we’re not getting along right now, and she
is leaving me out in the cold. And I just began to feel sorry for myself.

Etcetera. Then she goes on to say:

Well, it turned out that she said to me, when I said for Thanksgiving,
— 1 ’•“Well don’t I always have the family?” I said “Well you didn’t ask
me, how am I supposed to know what’s going to be this year?” I
mean generally I don’t stand on ceremony, but conditions are, they’ve
been different lately, you know.

A recurrent thing that I’ve seen throughout this stuff is persons talking
about not feeling wanted anymore. The question is, how is that kind of
feeling provided for in this society? And what would be interesting about it
would be if we could see some way in which, quote, the structure of society,
provided for the focussing of kinds of troubles. That’s what I think we can see

18

Part I

with this, 1 if we just consider ceremonials a little further. We can note that
there are classes of events which, between persons who are not terribly
intimate, get initiated via ceremonials. “Would you like to come over for
dinner tonight?” “Sure.” That is, for these kinds of events to occur, there has
to be an invitation, an offer of some sort. So that’s one task of ceremonials -
they do the job of providing for these events to take place.

They do another job, in a way. When persons are quite intimate, then one
way they measure that is by virtue of the fact that invitations are no longer
relevant. You can go over to their house without being invited. And people
will say to each other, “Come over any time you want.” Now with a husband
and wife, one gets a version of this not feeling wanted, which goes something
like this:

Wife : Why don’t you ever ask me to go out to dinner anymore?

Husband : If you want to go out to dinner why don’t you just say so?

Wife : I don’t want to go out, I just want you to ask me.

What she’s picking up here is the absence of ceremonials. And ceremonials
have this double use. On the one hand they are properly used to provide for
persons to do things - come over, go out to dinner, etc. - at some state of a
relationship. At another state those things happen without them. And they’re
not absent. Indeed, it surely happens that somebody might say, “Well why
don’t you come over tonight?” and the other says “Why are you suddenly
making a big deal of it?” But this double use then provides that when
somebody has some doubts of some sort, they could focus right in there; that

1 Throughout this volume many of Sacks’ pronominal uses have been changed.
Here, the operation is more or less innocuous. What is rendered as "... if we could
see . . .” and “That’s what I think we can see . . .” actually goes, "... if we could
see . . .” and “That’s what I think you can see,” i.e., the second ‘we’ is actually
‘you.’ This change instances an editorial policy concerned with solving ‘direct
address’ as a problem to a reader (e.g., “You ought by this time to be quite aware
of the fact that . . .”). The policy takes as a resource and license Sacks’ own use of the
pronouns ‘you’ and ‘we’ in alternation (e.g., “We want to do [X and Y], You want
a method that generates this.”), and his somewhat eccentric use of ‘we’ in particular.
For example, he will use ‘we’ when he himself is the obvious referent (e.g.,
“Remember we said about the opening lines of conversations that they seem to come
in pairs”) or when the class is the obvious referent (e.g., introducing a “much more
interesting thing that I doubt we’ve noticed”). For a more elaborated discussion, see
Appendix A.

Of the range of changes made to the unedited transcripts, very few are marked and
explicated. It might also be noted that the faithfulness of the unedited transcripts to
the very words is in principle suspect. Such preservation of the very words as there
is, is variable. That was not part of the enterprise - with the exception of the
retranscribed Fall 1964 lectures, which were produced after Sacks’ death. All of
which is to say that the spontaneous nature of the lectures themselves, the variable
fidelity of the transcripts, and the manifold changes made in the editing, result in a
version of Harvey Sacks’ work which from start to finish was in one sense or another
not under Sacks’ control.

Lecture 2

19

they see this thing is absent, and see the absence via the position of one who
is not in the position of intimacy. And they don’t know quite how to handle
that matter. Because if they complain, they get “Why are you standing on
ceremony?” and if they don’t complain and don’t get the invitations, they
figure “Jeez, it’s the case that I’m not wanted anymore.”

Now I can’t make any statement psychiatrically about why persons would
pick up that double use of ceremonials and use it - or feel used by it - with
their doubts. But in any event, one can see how it is that the fact that those
things get used that way, provides a locus for troubles to get focused on.

Here is another, related, phenomenon.

A: Hope you have a good time.

B: Why?

The “Why?” here is quite apparently a paranoid return, and the whole
conversation from which this comes makes it quite clear that the person who
produces it is paranoid. I won’t quote the whole conversation, I want to just
focus on this interchange for reasons I’ll make clear.

One of the things that’s reported about persons who have to deal with
paranoids is that they feel weak, experience a terrific lack of control when they
encounter them. Now you could go about trying to examine that, perhaps by
studying let’s say the comparative dynamics of the persons, or various other
things. But I think you can get an idea of how they would have that feeling
of weakness by just examining an interchange like this. We’re talking about
ceremonials. The normal answer to this “Hope you have a good time”is
“Thank you.” And if one uses a line like “Hope you have a good time” one
can expect to control the return of the other. In this case the line doesn’t
control the return of the other, and we can at least begin to see what it means
to feel weak: Having an expectation of doing something as controlling, and
finding out that it isn’t that at all.

But furthermore, this “Why?”-return casts this “Hope you have a good
time’ ’ into the character of an ‘accountable act. ’ Normally, when one does an
accountable act, one knows that one is doing an accountable act. This one
comes off like this:

A: Hope you have a good time.

B: Why?

A: Why? Well, I just would like- you know, you ought to have a good

time if you’re going on a trip.

What seems to be involved here is, doing something that wasn’t seen as
accountable, having it turned into something that is accountable, one doesn’t
have an account. One offers, then, an account which one feels is quite feeble.
It’s feeble in a special sense: Not only is it inadequate, but it’s inadequate by
virtue of the fact that there’s no reason to have had an account in the first

20

Part I

place. But when one delivers the account, one may only see that it’s feeble,
and get the sense that, “Jesus I’m behaving inadequately here.”

And that character, that others can by virtue of their return cast your
activity into something other than it was produced to be - or that they can by
virtue of their return cast it into what you thought it was - is a very basic
problem. I call it Job’s Problem. Remember the Book of Job? Job is a rich
man, doing marvelously. Then everything is destroyed. Job’s position is that
he didn’t do anything wrong; this was not ‘punishment.’ And now his friends
come, and they say to him, “Just confess to what you did wrong and
everything will be fixed up.” That is to say, the appearance of his pain and of
his loss is sufficient indication for them that he did something wrong. And the
problem as they see it is that he isn’t about to confess to it. Job, then, is in this
position of, “Well Christ, the world has changed for me. And maybe I did do
something wrong.” But he is not about to acknowledge that. But most people
do. Most people, when they get into a situation, will say, “What did I do
wrong?” or “What did I do to deserve this?” That is to say, treatments are
‘proper treatments.’ And one isn’t in a position of saying right off, “You’re
treating me wrong.” Rather, one finds, the treatment occurred and it must be
about my action.

Lecture 3

The Correction— Invitation Device

Let me give a quotation.

A: Do you have a gun at home?

B : A forty-five

A: You do have a forty-five.

B: Uh huh, loaded.

A: What’s it doing there. Whose is it.

B: It’s sitting there.

A: Is it yours?

B: It’s Dave’s.

A: It’s your husband’s huh

B: I know how to shoot it.

A: He isn’t a police officer.

B: No.

A: He just has one.

B\ Everyone does, don’t they?

A: You have a forty-five and it’s loaded.

B: Uh huh,

A: And I suppose maybe everyone in Burnside Park has one. I don’t

know.

B: No. But I mean a lot of people have guns. It’s not unusual.

A: Oh sure. I see.

The first thing I want us to see in this, I think we have two of, more or less.
It’s this use of “Is it yours?” and then this one, “He isn’t a police officer.” I
want to call them, and things like them, ‘correction-invitation devices.’

By that I mean: Where one wants to get, from the person one is talking to,
an account of something - why they did something or why they have
something - one way you can do it is by saying “Why?” Another way you can
do it is by asking with the name of the class of things you want. For example,
a woman is talking to an officer from the juvenile division of the police force.
Her 14-year-old daughter hasn’t been coming home at night. The woman
called the police, the police found the daughter, and now they’re talking to
the woman. And they say, “Have you ever had this kind of trouble with her?”
That is, ‘this kind of trouble’ is the name of the class. She can then say, “No

Tape 5, side 2, and a brief extract from Winter 1965, lecture 7 pp. 12-13
(transcriber unknown).

21

22

Part I

I haven’t had this kind of trouble,” she can say “Yes” and then give some
instances, or she can say ‘‘No I’ve had other kinds of trouble.”

Now it also seems that one can ask for an account by naming, in question
form, one member of the class, of which the account will be another member.
For example, “Is it yours?” She doesn’t come back and say just “No,” though
people sometimes do that. She says “It’s Dave’s.” That is, instead of saying
“Whose is it?” which he said earlier but didn’t get an answer to, he gives one
possibility and thereby elicits, as its correction, another; the actual class
member.

For “Fie isn’t a police officer,” the problem is, how is it that the husband
happens to have a gun? There are classes of good accounts which would
explain why somebody has a gun - that is, has a gun properly. One member
of that class is ‘police officer.’ And what could happen is, if “He isn’t a police
officer” is an instance of the correction-invitation device, and if the device had
‘worked,’ then the return would be, “No, he’s a such-and-such,” or “No, we
have it because ...” Here’s another example. Two persons are talking on the
phone:

A: What do you think was the cause?

B: It’s a little difficult for me to speak now.

A: Oh it is. You’re feeling badly yourself?

B : Oh no. It isn’t it. I’m lacking in privacy.

A: Oh you’re lacking in privacy. Well, why don’t we arrange to talk

tomorrow.

“You’re feeling badly yourself?” would be one account of how it is that B
finds it a little difficult to speak now, and the return is the correct account.

I’ll just mention one way that these things get used, which can get us to one
basis for their use in the first place. When police interrogate persons, one thing
they do is, instead of saying “Are you the guy that murdered this fellow?”
they say “Did you hit him with a tire iron?” And the guy says “No,” and
then they say “Well what did you hit him with?” where the guy hasn’t
admitted yet that he did it. And it may be the fact that this form is so
routinely used elsewhere that permits it to set up the possibility of a trap like
that.

Now, so far I’ve talked about the construction of these correction-
invitation devices, and said that it’s based on the fact that, using a range of
classes, you can refer to one member to get another member. We might also
be able to say something about the basis for their being used in the first place.
And at least one basis for that is perhaps something like the following. If you
say to somebody “Why did you do this?” then what they are being asked to
present is something they may well know they have to defend. And you set
up a different situation when what they have to present is something they
know they have to defend, as compared to setting it up such that you’re
not asking for an account they have to defend, but you’re ‘inviting a
correction.’

Lecture 3

23

If these different forms can set up basically different situations, that would
suggest that we’re looking at extremely powerful matters. I don’t know that
they’re that powerful, but if they are, that’s a very important thing to know.
That is, that just by the way you set up the matter, without regard now to the
consequences in a large sense - as in the murder interrogations - this thing can
work.

The fact that these things are not only recurrent, but that they do work,
makes them worth looking at for the following sort of reason. Sociologists
often talk about something called ‘common knowledge.’ And one question is,
what is it that common knowledge consists of? One thing it can consist of is
just lists of items that persons know in common. But there are some things it
would be nice to know about the phenonenon of common knowledge. One
of them is what we could call its ‘structural properties’ — and we’ll talk about
lots of them, I hope. Also, how it is that what persons know ‘in common’ is
organized. Also, is it the case that the organizational features of what they
know ‘in common’ are also known?

So if persons know that there are classes of accounts for some action, the
question is how do they know those classes? For example, do they know them
only if you name the class, then they know one or another which are members
of the class? What this stuff seems to suggest is that on the one hand they
probably do know, to some extent, the classes and items of these classes by
virtue of the class name; ‘kinds of troubles,’ for example. But they also know
them in this fashion: You can name one, and they know, by virtue of the use
of that one, what class you’re referring to, and can give you another. And
that’s a non-trivial way of seeing that, and how, common knowledge has its
organization seen and understood by Members.

Now I haven’t yet been able to track down when this thing works and
when it doesn’t work, or what we might say about the circumstances where
it might be clear that a person knows how to use this, and knows what
another account is, and doesn’t pick it but instead answers only ‘the question:’
“He isn’t a police officer,’’ “No.”

And in that regard, another question about this organization of common
knowledge and the members of the class ‘accounts’ would be, how substi-
tutable are accounts for each other? Is it the case that one is as good as another?
Which ones would be as good as another for this or that account problem?
And I ought to mention that the correction— invitation device may not only
work for accounts, but for all sorts of things; that is, where you can name an
item, and get in return another item.

Let me turn to another sort of thing that we can see in this piece of
conversation. This line, “Everyone does, don’t they?” is one of the most
fabulous things I’ve ever seen. Where persons are engaged in trying to get an
account from somebody, there’s an object that the person who’s being
questioned can slip in. This is one of them. And what it does is, it cuts off the
basis for the search for an account. I don’t have a terribly elegant name for it.
What I called it was, ‘account apparently appropriate, negativer.’ Or A3N.
So, for example, having a gun is something for which an account is apparently

24

Part 1

appropriate. The search goes on for an account, this thing goes in, and now
an account is no longer to be sought. And this thing isn’t an account of how
she happens to have a gun. That would be quite a different thing.

Now these are extraordinarily interesting. One of the most interesting
things about this one in particular - though it’s not so for all of them - is that
it seems to be a ‘general purpose A3N.’ By that I mean, it doesn’t much
matter what it is that you’re seeking an account about, you can use this one,
‘everybody does.’ This object cuts off accounts about God knows what —
where accounts are, of course, extremely crucial phenomena. It’s a general
purpose device. And we’ll see some more later on, some of them much more
extravagant than this. Just consider, with respect to the organization of the
social world, that we’re told how fantastically complex it is. How everything
is a blooming, buzzing confusion. How everybody is different. Etcetera. That
there are these general purpose devices might give a glimmer, perhaps, of an
extraordinary kind of simplicity.

By and large I’ve only talked about verbal interaction. Let me just mention
something that isn’t a verbal device. My parents live in an ‘exclusive’ suburb.
And when I was a kid in high school I always used to walk around at night
in the streets. And when you walk in the streets at night in exclusive suburbs,
you’re liable to get - as I was routinely - picked up by the cops. “What are
you doing?” “Just walking.” Then they would take me and stand me in front
of the police car with the light shining in my face and call up the police station
to find out if I indeed did live there. This happened night after night. Finally,
someone gave me the solution. If you bought a dog, that was the end. You
never got stopped. And that has now become a matter of common
knowledge. It has become so much a matter of common knowledge that in
the book Beverly Hills is my Beat by a Captain Anderson, head of the Beverly
Hills police force, he writes, “It used to be the case that an excuse to walk the
streets was having a dog. However, the robbers started walking around with
dogs. But don’t try it in Beverly Hills, because we also know the dogs.”

So we can begin to locate a range of general purpose A3Ns, with greater
or lesser generality of application. ‘Everyone does’ has enormous generality.
Another thing to notice about it is, it doesn’t seem that evidence needs to be
offered. That is, it’s not the sort of statement about which someone will say,
“How do you know that’s so?,” where there are lots of statements which will
get such a question. It’s been known for a long time that there are classes of
objects - a very predominant class of which are proverbs - about which, on
the one hand, Members don’t have doubts, and on the other, it’s not a matter
of evidence that they’re so. And the existence and use of such objects is fairly
obviously the basis for a great deal of philosophy. Hume, for example, talked
about the fact that when he was sitting and doing philosophy, there were lots
of things he could doubt. But he found that as soon as he got up and walked
out of his study, they were just there. And in an important sense, he had never
doubted it. It may well be that these are the sorts of things he was trying to
figure out what in the world they were, and how it is that they seem to do
what they do.

Lecture 3

25

Again and again we find that when such general-purpose devices as A3Ns
and proverbs are used, others don’t attempt to question them or contradict
them. I think there’s some reason why we don’t much see attempts to
question or contradict these things, and that is that they may be such basic
objects - that is to say, Members are so committed to their correctness - that
if you undercut one, exactly what you’ve undercut is not clear. And one
doesn’t know exactly how we can continue talking.

My reason for saying that is the following. A woman was collecting
research materials by going into parks with her children and just starting
conversations with people. One of the things she reported was how the
conversations began. And one recurrent way they began was, there would be
a woman sitting on a bench. This woman would go over to the bench with
one of her children, and sit down. The little boy would wander around for
awhile, then he’d come up to her and she’d say, “Go away, I want to sit and
rest.’’ Sometimes he’d go away, but sometimes he’d sit there, annoyingly.
And then the other woman would turn to her and say, “They’re all like that,
aren’t they.” And she’d say “Yeah” and they’d get into a conversation. I
asked her, “Did you ever say no, or something like that?” And she said
“Yeah, when I first got out of college I was all full of information. People
would say that to me and I’d say ‘Well I don’t know, my kids aren’t.’ And
they’d always stop talking right then and there.”

I don’t know that that’s generally true. I’d like to see whether it’s so, that
if you don’t express commitment to these sorts of things, person feel that they
can’t really talk to you. But apparently proverbs and things like them have to
be affirmed or membership is not seen as something both of us hold in
common. That is, these things are known to be so - whatever that means -
but if you ask for evidence for them, then apparently it’s not clear what kind
of a box you’re opening up, what sort of things you’re going to ask for
evidence about next, what would stand as acceptable evidence. So they’re just
known to be so, whatever that may mean, in that they can be asserted, they
can be used in conversations, etc. They’re not known by virtue of ever having
been established in some specificable way. In that sense, they’re strictly
traditional pieces of information.

Lecture 4

An Impromptu Survey of the
Literature

Books like Plainville, Street Comer Society , The Gang, The Irish Countryman,
and a series of others, were part of a kind of sociology done in the United
States mainly 25 to 30 years ago. It’s associated with the University of
Chicago, and also with Harvard. At that time those fellows were trying to
build ethnographic studies in a tradition that had been developed largely in
England in social anthropology, and there largely by studying tribal societies.

That work essentially died out in the United States. But in recent years,
anthropologists are again returning to detailed ethnographic work, and the
term ‘ethnographer’ which had fallen into considerable disrepute, has been
adopted as an ‘in’ term. The Urban Villagers by Gans is one recent book
which is again attempting to do that sort of work. Two other recent books in
the same vein are Millways of Kent by John Kenneth Morland and Blackways
of Kent by Harlan Lewis.

This recent work is of a new sort, in a way. Where much of the early work
was criticized as being impressionistic, casual, not hard; that is, not repro-
ducible, not stating hypotheses, etc., the new ethnographic work - which is
calling itself things like ‘ethno-cognitive studies,’ ‘ethnocultural studies,’
‘ethnoscience’ and the like - is attempting to proceed without being subject
to those criticisms. The concern is to try to describe the categories that
members of a society use, but to describe those in a very hard fashion.

There are several bases for this renaissance - if it’s a renaissance. First, the
development of very strong tools by linguists. Second, the impact of the work
of Whorf, whose collected papers are now in a paperback called Language,
Thought and Reality. Third, and of pretty much equal importance, is
Wittgenstein’s later philosophy, which the anthropologists who do this work
are familiar with. The relevant books are two volumes put out as one called
The Blue and the Brown Books, of which The Brown Book is the easiest
introduction, though it’s not easy, and the book called Philosophical Investi-
gations.

Tape 3. These materials were probably produced in response to a student’s
question.

Transcriber unknown. This is the first of several lectures comprised wholely by an
early transcript for which there is no tape; i.e., which could not be retranscribed.
There are several of these ‘unknown’ transcribers whose work is included here,
without whom the first set of lectures would be significantly impoverished.

26

Lecture 4

27

My own relation to that stuff is fairly tangential in some ways. Instead of
pushing aside the older ethnographic work in sociology, I would treat it as the
only work worth criticizing in sociology; where criticizing is giving some
dignity to something. So, for example, the relevance of the works of the
Chicago sociologists is that they do contain a lot of information about this and
that. And this-and-that is what the world is made up of. The difference
between that work and what I’m trying to do is, I’m trying to develop a
sociology where the reader has as much information as the author, and can
reproduce the analysis. If you ever read a biological paper it will say, for
example, “I used such-and-such which I bought at Joe’s drugstore.” And
they tell you just what they do, and you can pick it up and see whether it
holds. You can re-do the observations. Here, I’m showing my materials and
others can analyze them as well, and it’s much more concrete than the
Chicago stuff tended to be.

And I differ from the modern anthropologists, though I would recommend
that work very much. There is a paper by Hymes called ‘The ethnography of
speaking’ which is quite interesting. And then there is a collection of these
anthropologists' most recent works called Contributions to Cultural Anthro-
pology edited by Ward Goodenough. The trouble with their work is that
they’re using informants; that is, they’re asking questions of their subjects.
That means that they’re studying the categories that Members use, to be sure,
except at this point they are not investigating their categories by attempting
to find them in the activities in which they’re employed. And that, of course,
is what I’m attempting to do.

There are other matters of a deeper sort which are perhaps relevant to why
sociology took the course it did, and they’re intrinsic in Durkheim’s work.
One of them is the notion that the order of social events is macroscopic, in the
sense that you had to assemble lots of events to find statistically what it was
that was doing the work. I think one can begin to see, in the stuff I’ve been
talking about, that it may well be that things are very closely ordered. And
what we have may be something like the following. There may be collections
of social objects - like “How are you feeling?” - which persons assemble to
do their activities. And how they assemble those activities is describable with
respect to any one of them they happen to do. That’s a different kind of order
to a social world.

In a way, most of sociology could have been irrelevant, and what I do could
have been done 50 years ago, 100 years from now, etc. As I said before, it
stands in close parallel to classical naturalistic biology or zoology. 1 In fact, if
you want to look at something, Darwin was doing this already. He wrote a
book, Expression of the Emotions in Man and the Animals , where he collected
pictures and tried to see if there were, for example, standardized expressions,
and if so, how did they operate? Not until 50 to 60 years after that book was
published did people again begin to use photographs. So, for example,

1 This may be a reference to an introductory lecture, of which we have only a
paragraph-long precis by an unknown transcriber.

28

Part I

Bateson and Meade began to do that some 25 years ago in Balinese
Character.

But by and large the direct study of humans and their behavior wasn’t
done. And it probably wasn’t done because nobody believed that it could be
done, or perhaps because it wasn’t interesting for some reason or another.
More recently, those who have tried to study it very closely - for example,
Bales in his laboratory work - have done something exceedingly foolish, I
think. That is, Bales has the notion that you can categorize it as it comes out,
so that you sit and watch people as they are talking, and write down categories
of what they’re doing as they’re doing it. That makes it into some kind of
trick. There’s no reason to suppose that you should be able to see it right then
and there. (I find it hard to imagine, for example, that a fellow would stand
next to an electroencephalograph machine, or any other such machine, and try
to give you an analysis of the data as it comes out on the tape.) Instead, you
take these little pieces and you try to collect those that look alike, and it can
take an awfully long time to understand any given one.

Another thing that might have been involved was the notion that you
could tell right off whether something was important. So you would start to
look at what kings did, or to look at votes, or revolution, for example, because
those were obviously important. But, for example, the whole of biology has
been revolutionized by the study of one bacteria, though when that bacteria
was first being examined, no one had any idea that it would do that work.
And it’s possible that some object, for example, proverbs, may give an
enormous understanding of the way humans do things and the kinds of
objects they use to construct and order their affairs. That has to be seen by
attempting to analyze the stuff.

And in that regard, a debt is owed to Freud, who did say “Now let’s treat
patients as sacred phenomena. ’ ’ That is, something that you would study in
the sort of way that biblical critics have studied the Bible, where the fact that
you were looking at one line wouldn’t mean that you could only write a page
on it. You could write 100 pages. You could spend your life studying it. The
reason that sociologists haven’t studied a line is that they treat it as something
very ephemeral, where if you treat it as a machine itself, and as enormously
recurrent, it has quite a different character. But, for example, the American
philosopher Meade was a most extraordinary figure who proposed that
psychology was the study of that which is not available to observation. He
had an enormous impact on sociology, God knows how or why. It may in
part have had to do with the notion that sociology studies ‘society,’ which has
been a very troublesome idea because then you start out by saying, “Well,
society isn’t observable, but Meade has shown that you can study things
which aren’t observable. So let’s go study things which aren’t observable, like
attitudes.”

But social activities are observable; you can see them all around you, and
you can write them down. The tape recorder is important, but a lot of this
could be done without a tape recorder. If you think you can see it, that means
we can build an observational study, and we can build a natural study.

Lecture 4

29

Now all of this is background. I don’t want to go through the history of
sociology and show why one does this or that, because first of all if you want
to do it seriously, you have to know what kind of work theorizing is, and that
is an extremely obscure domain if you’re going to take it seriously, at least as
I take this stuff seriously. I have no idea why sociologists do what they do, and
I don’t want to get into long arguments about matters which really can’t be
taken seriously. My arguments can’t be taken seriously, Mills’ arguments
about the effect of Parson’s proposal to reraise the issue of ‘are ideas important
and what kind of resources do we have for asking that question’ can’t be
taken seriously. We can talk about it as philosophers in conversation, but to
talk about it in any serious way presumes that we have an enormous amount
of information about how the animal operates, which we don’t have. So the
more material you have at your command, the more you ought to be able to
pick up items and see their recurrence and get some idea of what they might
be doing. But the way to proceed is item by item.

Q: How does this fit into the general definition of sociology and social
structure, as structured?

HS : That’s a good question, for the following reason. What has definition
got to do with anything? Let’s consider what a definition can do. A definition
could be an epitaph to be put on a headstone: ‘That’s what this was.’ The
notion that it’s a control of activity - that is, if you don’t define what you’re
saying you can’t do anything - is an absurdity. Just consider, for example, the
fact that biology was said to be ‘science of life.’ Now they find out that maybe
there wasn’t anything such as that, or that it doesn’t make much difference.
In any case, it never controls any of their work. So one doesn’t know
what ‘life’ means until one knows all the things that the biologists try to
show. And you may find out that it’s not the kind of thing you started out
with at all.

There isn’t an answer to what you ask. But I’m not sure what you asked
had, in any deep sense, a right to be asked. If you take it, as I do, that a
question has to arise out of something you’re trying to deal with, as compared
to a way you’ve been taught about questions, then your question might
not arise.

It’s a big problem about the University as a scene, that you have almost a
free right to ask questions. You can turn almost anything into a question, and
it’s not insane. You learn that much of the knowledge that you’re going to get
is formable as the answers to questions, because after all you have to be given
it in such a way that you can answer it if I use it as a question on an exam.

Now that’s a fantastic constraint on scientific research - that, for example,
the product of research is subject to being used in a quiz. My own feeling is
that it was the death of academic psychology that it grew up in a university.
That implies that they did experiments for which it could be seen from the
start how the results of those experiments would look as answers to quiz
questions. If a research couldn’t be farmed as a quiz question eventually, then
maybe you just didn’t do it. Other fields, like biology and astronomy,
developed well outside of the universities, so they had lots of materials already

30

Part I

worked up by the time they got into the universities to be taught in this
fashion.

And there is, for example, the notion of ‘courses,’ where there’s a
beginning and an end - an organization. What has that to do with the
physical or social phenomenon you’re studying? And textbooks. You have to
have introductory textbooks in sociology or sociologists don’t know anything.
And the way textbooks teach sociology is quite exquisite. I’ll give a marvelous
example of how you come to learn sociology. There’s a line in Broom and
Selznick that goes like this: “Roles are more complex than they appear to be
at first glance.’’ Now there’s a basic sentence that you know as Members
without having done any sociology, which goes: “X are more complex than
they appear to be at first glance.’’ And ‘roles,’ which first of all is a concept
that couldn’t even be looked at ‘at first glance,’ now becomes something we
learn via that basic sentence that provides a blank for it. And you know a lot
about roles, you think. A book like this, built up out of these basic ways that
you already understand your world as a Member, and simply fits into slots,
is an especially powerful introductory textbook. But what you’re learning is a
batch of terms, which you can figure you now know something about, by way
of what you already know about everything that could fit into that sentence.
But for us, the understanding and use of objects like ‘X are more complex
than they appear to be at first glance’ is precisely what we want to be
studying. It’s not something that we can employ to give us the feeling that we
understand what is going on in the first place.

Q: How can you repeat the recognition unless there is some label to
communicate a definition?

HS : How do children learn to see ‘another bird’ when they saw a bird once,
or to re-see a car or a friend? It’s a very obscure question, how it is that persons
learn to see generalities, or see objects again, or see ‘another’ of an object. I
really don’t know how they do it, though it’s an important thing to learn how
they do it. But they do it, and they do it with all sorts of things. You do it
with verbs, adjectives, sentences; for example, you can see “There’s another
sentence.” You look at them just as objects.

Now, what we’re doing is developing another grammar. Right now I’m
using it with respect to verbal activities, or things like gestures, etc. And in the
same way you don’t have any trouble seeing a variety of things such as birds,
or “There’s another verb,” you leam to see these things - at least people come
back and tell me all the time, “Oh, there’s one of those things you
mentioned.’ You can see it working, doing the thing it does. We want to
name these objects and see how they work, as we know how verbs work, and
that sort of thing. We want to see how activities get assembled, as with a verb
and a predicate and with whatever else, you assemble a sentence. The category
that you use to name that activity is given by the Members. They have these
category names, by and large. So what you’re after is a way of describing the
activity that they have a name for.

Ideally, of course, we would be producing formal descriptions, as you
could give a formal description of how you assemble a sentence. It will not

Lecture 4

31

only handle sentences in general, but it will handle particular sentences.
Grammar, of course, is the model of closely ordered, routinely observable
social activities.

“This is Mr Smith, may I help you?” “Yes, this is Mr Brown” was one
kind of object and worked one kind of way. The way it works is the essential
thing about it. There may be a large amount of things that Members can do
with these objects. Or there may not be a large amount of things because
someone might do something and people will laugh at them, put them away,
etc. There may be lots of things that people might never do with these objects.
What we’re interested in is what do they do with them? Whether that’s
indefinite, definite, or not, is an empirical question.

We can say some of the things they may do with them. They may do other
things, they may not do other things. For example, “This is Mr Smith, may
I help you?” I have not seen occurring in the middle of a conversation, though
“Hello” does occur in the middle of a conversation. For example, when you
say “Wait I have to do something” in the middle of a telephone conversation,
you may come back and say “Hello” though you’d been talking for ten
minutes. So the latter has a use that the former doesn’t seem to have. “This
is Mr Smith, may I help you?” is strictly a greeting. The big thing is to see,
what are the properties of an object which permit it to do this or that task? For
example, the way the ‘repetition device’ works. There are special properties
which provide for its use.

If you really want me to talk about what sociology ought to be about, or
what relation any of these things has to what I do, I wouldn’t want to do it
in class, because that’s like taking a position. These can’t be handled seriously
unless one takes them as the kinds of issues they are; like take a line out of
a book and try to see how that fellow came to write that. And who knows?
At least I don’t know.

Lecture 5

Suicide as a Device for Discovering
if Anybody Cares

I’ll begin with a quotation. This is a suicidal woman, 40 years old, divorced,
no children.

A: Well perhaps you want to tell me uh why you feel like committing

suicide.

B: ((sigh)) ((sigh)) Well it’s the same old childish reason that everybody

wants to commit suicide.

A: Why is that.

B: You want to find out if anybody really does care.

There’s a whole bunch of things that are interesting here, and large collections
of things we have to do if we were going to be able to generate this
interchange, most of which I’m not going to consider now. For example, you
might look at the way this caller sets up giving her answer - by the use of
“Well it’s the same old childish reason that everybody wants to commit
suicide’’ - and compare it to the A3N device. 1 That is, the A3N can provide
that an account need not be produced. The sort of line this woman uses might
provide that the account she is about to produce is not challengeable, needs
no defending.

We might also notice how that’s added to by the use of “you.” That is,
instead of saying “I want to find out if anybody really does care,” she says
“You want to find out ...” And those usages, where a person says “you” or
“one” as a way of stating something that they propose thereby to be a
generally correct remark, and how they are defended, and what kind of
attacks they can be subjected to, are something we can watch. And I’ll deal
with these matters later on.

I’m now mainly concerned with “. . . if anybody really does care” and not
the particular objects by which this sequence gets done. But I do want to note
the fact that this first question, “. . . tell me why you feel like committing
suicide,” can be asked as a sensible and appropriate question to which there
is expectably or reasonably an answer - that why you want to commit suicide
is something that you would have information on, or could propose to know.

A combination of Fall 1964, tape 4, side 2 and another lecture, ca. Spring 1965
(‘64— ’65).

1 See Fall 1964-Spring 1965, lecture 3, pp. 23-4, above.

32

Lecture 5

33

That the question is askable can be considered this way: Given that there are
sets of question forms which Members use, one of which is “Why do you
want to do X?” where ‘X’ is some activity, then, ‘suicide’ being an
activity-category, just by reference to the relevance of that form for any
activity, it can be applied to suicide.

How it is that such a question can expectably or reasonably be answered is
worth some consideration, since for professionals there are classes of things
which, if you do them or want to do them, then ipso facto you don’t know
why. And psychiatrists — and psychoanalysts in particular - take it that a
person who wants to commit suicide doesn’t know why they want to commit
suicide, in the sense that the psychiatrist could say why they want to commit
suicide. (And of course sometimes a person says “I don’t know.”) Now that
fact doesn’t seem to stand in the way of asking the question. And the issue
then is, what’s the relevance of that question, and what would happen, insofar
as persons come to know what it is they didn’t know? That’s Socrates’ classic
problem; that one thing about knowledge is that you know what you don’t
know, and to the question “Why?” the answer “I don’t know” is sort of a
deeper answer; that is, it might have an awareness of the character of this
knowledge as something only professionals have.

The notion of ‘opinion’ as contrasted to knowledge (and Plato made a great
deal of the difference between them) and the sheer introduction of a notion of
‘opinion,’ provides in part for professionals’ talk to laymen. Because one of the
characteristics of ‘opinion’ is that it’s something which lay persons are entitled
to have when they’re not entitled to have knowledge — in the sense that they
can offer it without ever proposing to have to then defend it. Like they say “My
feeling is such-and-such on that, but I don’t really know,” as a permissable
way of talking, where one then doesn’t try to find out what kind of defense you
have for that statement. So in a way, ‘opinion’ provides for the continuing
discourse between professionals and laymen. And I presume that it’s a means
or a mechanism by which not just psychiatrists, but perhaps professionals in
general can talk to clients — by the notion of the permission that ‘opinion’ gives
to a person to talk. That is, under the control that one doesn’t really know;
which is to say, one isn’t entitled to know. And very frequently when you see
“I don’t know” appended to some statement, that’s what it seems to be doing
- providing that “I’m not entitled to say this,” that is to say, “I can’t defend
it professionally,” if it’s a matter of professional information.

But if it’s the case that there’s going to be discourse between clients and
professionals, or between the public and professionals, then the fact of a
distribution of knowledge which provides that professionals know and
laymen don’t know might seem tremendously interruptive unless you had
some mediating device, like ‘opinion,’ which would permit laymen to keep
talking even when they find out that they don’t know. Otherwise they might
not have any way, for example, of even turning to a professional.

What I want to focus on is, why is it that suicide seems to be a way to find
out if anybody does care? The question I asked when I was sitting trying to
puzzle that out was, what are the available ways in this society for going about

34

Part 1

determining that others care, or that one is relevant to others? What are the
means available for seeing one’s relevance?

And while I had that stored at the back of my head, I was reading one of
the greatest books in the social sciences, Witchcraft, Oracles, and Magic Among
the Azande by Evans-Pritchard. And some of his observations can begin to
give us a feel for what such a procedure might look like. Here’s what he
reports. Whenever anything goes wrong among the Azande - if an Azande
feels lousy, gets sick, injures himself, is economically in trouble, etc. - he
engages in the following procedure. He pretty much drops whatever he’s
doing and goes off into the woods with some oracle procedure. Like, say, one
oracle procedure is they take a chicken and give it a little poison and ask
questions to the chicken, which the chicken answers by dying or not dying
upon being given the poison.

So the Azande takes a chicken and some poison and goes off into the woods
with it. And he sits down and makes up a list, essentially composed of his
neighbors. He considers what his state was before he got ill, and then goes
through this list of neighbors, considering about each person how he takes it
they feel about his situation. Are they unhappy that he just got married that
week, that he just got some wealth, etc.? By going through this procedure he
then locates some persons who he figures would like to cause him trouble.
And for each person that he has in this way, he offers a name to the chicken
and gives it some poison. On some giving of poison the chicken will die. The
person whose name was offered on that occasion is the person who has done
him the trouble - caused him to have some illness, caused the rain to fall
before his crops were in, caused him to have a bad hunting trip, etc., etc. And
once the one who caused the trouble is found out, there is some procedure for
getting amends.

Evans-Pritchard reports that the Azande just love to do this. There is pretty
much nothing that will stand in the way of them stopping and going off into
the woods and making up a list and sitting down and considering, for all the
people around, ‘How are they interested in my good or bad circumstances?’
Now, this is one rather nice kind of procedure, which is institutionalized in a
society, whereby persons can take an occasion and determine for themselves
properly - that is, there is a proper occasion for doing it - whether anybody
cares, and what they care.

Let me make a parenthetical remark about the situation of the Azande as
compared with this society. One of the things that lies at the basis of the
availability of that procedure for the Azande, and which is not present in this
society - and which then provides that we don’t do that in this society — can
be stated in the following way. The Azande do not have an institutionalized
notion of chance. Things like falling ill, and most particularly things like
dying, do not occur by chance for the Azande. There is always somebody
who’s responsible. And there is a set of procedures, the purpose of which is
to find out who it is that’s responsible. And these are not random procedures,
because one has some way of finding out, in the first place, who would be
interested.

Lecture 5

35

Now it’s not that the Azande don’t have a good notion of ‘natural causes.’
They are perfectly well aware of of the fact that you can get ill from natural
causes. That doesn’t exclude the fact that there’s somebody interested in those
natural causes occurring. Evans-Pritchard reports, for example, that some-
body will stub their toe on a tree and then go off with their chicken.
Evans-Pritchard says to the guy, “Well after all, you know, it’s your fault.
You stubbed your toe on the tree.” And the guy says, “I know perfectly well
that I stubbed my toe on the tree, and that the tree caused that trouble, but
Eve been through this forest hundreds of times and I never stubbed my toe
before. There must have been some reason, then, why it happened this time.”
And that, then, provides for the responsibility. So it’s not a matter of they
don’t have a good notion of natural causes. It’s that they don’t use a notion
of chance.

That being so, you can come to see how rather special it must be for a
notion of chance to be in fact enforced, and how easy it might be for it to
break down. Because what a notion of chance involves is that something that
happens to you is not a matter of inquiry as to how it came about. It just
happened. You simply don’t investigate why this or that trouble arises, for a
great many troubles. And that might provide for people to do you ill in more
or less subtle ways. The notion of chance is a pretty tender one anyway, and
persons suffering various troubles in our society will often feel that they have
to shed it and begin to employ, for any given trouble, the question “Who did
that and why? What do they have against me?” That is to say, they no longer
feel able to - or they feel compelled not to - use a notion of chance where
others use it. But in this society it’s not proper, and in fact it’s diagnostically
significant, if you do not use the notion of chance. By ‘diagnostically
significant’ I mean persons who do not have a notion of chance are persons
who have the symptoms of paranoid schizophrenia. When some trouble
befalls them, they take it that it is some persons who are in the business of
generating it for them.

Okay, end of parenthesis. For the Azande, then, there is a device which is
routinely employable for checking out how it is that others attend to your ill-
or well-being. Once we have some idea what such a procedure looks like, we
can begin to consider what are the sorts of things that look like that in this
society? What are the occasions under which one can make up a list like that
and just sit down and consider who cares and what do they care? I think you
can find that there are very, very few.

One such occasion is the wedding. Before a wedding the parents of the
bride sit down with a big list and have this enormous ball considering
“Would this fellow be happy that our daughter is getting married?” “How
would this guy feel?” Some people give parties, they say, to occasion such a
device; that is, they say “I just gave a party to see who my friends are.”

But I take it that the most prominent occasion in, so to speak, a person’s
life, is right after they die. In this society, on the occasion of death, people
gather around and talk about how important so-and-so was to our lives, how
much we cared about him, how much we miss him, what a marvelous guy

36 Part I

this was. And that’s what this suicidal woman reports. Later in the call she
says:

B : And daddy died he won’t suffer anymore now the family won’t be

aggravated And he’s not here aggravating other people He was
aggravating everybody before he died and as soon as he died you
know he wasn’t aggravating anybody anymore so they just said he
was a great guy

And anybody who’s ever witnessed that scene has learned what an opportu-
nity it is. And of course it’s a well-known fantasy, seeing yourself as the one
who died, getting a chance to get those credits which persons never give you
and that you can’t yourself collect - that is, for which there’s no occasion to
collect them. You can see how, for someone in pain, that scene after death —
which is known to everybody as an occasion for having persons propose that
they care about somebody - may then come as something exceedingly
attractive, and ‘the only way.’ And how, then, the ‘attempted suicide’ can be
the attempt to actualize that scene.

There are, of course, less dramatic devices for considering somebody’s
relevance by reference to missing them, or absences. For example, when
somebody comes back from somewhere, the question is, “Did you miss
me?” as a way of deciding whether it is that one cares. The question of
absence and loss, then, seems to be a basic way that one has of dealing with
relevance.

Now there are other, more specialized devices for doing a similar task. I’ll
start considering one of them in a slightly tangential way. One of the things
I came across several times in the telephone conversations I’ve been analyzing,
involved a widow or widower who was suicidal. They would say that time
hangs heavy on their hands and what they find is that “nothing happens.”
Nothing happens to them. And I wanted to see if there was some way of
finding out how that comes about; that somebody sees that nothing happens
to them.

I also have conversations between young married persons. And one of the
most exquisite kinds of things that young married persons do with each other
is, they say things like, “Kennedy was assassinated two weeks after we got
engaged.” I want to give the name ‘private calendars’ to that sort of talk. And
I want to note that married couples, each one, by themselves, independently,
construct these private calendars. And what private calendars do is to provide
for the locating of, not only events within that relationship, but events of the
world in general, by reference to the relationship.

Further, these calendars are ‘causally powerful.’ What I mean by that is,
there are all kinds of events which can be explained by reference to the
relationship. There is a generic statement: ‘Because A did X, B did Y,’ where
one can substitute for A, ‘wife’ and for B, ‘husband,’ and substitute for Y the
event to be explained, and for X the activity which can explain Y. This
provides a large class of sensible statements which persons in units like

Lecture 5

37

husband-wife are able to employ. Indeed for many events, such statements
have to be employed; that is, for many events, such an explanation is the
only sensible explanation. So it’s often said that while you can give a whole
list of explanations for why it is that somebody succeeded, in the last analysis
it’s because of his wife. It’s said without knowing the guy, or knowing
anything else.

Another sense in which the private calendar is causally powerful can be seen
in the paradigmatic statement, “That was before I met you and I was lonely
then.’’ There is a class of logical statements which the logician Nelson
Goodman named, and pointed to as creating very basic problems for the
philosophy and logic of science. He calls them ‘counterfactual conditionals,’ of
which an example is, I think, “If one had lowered the temperature to
such-and-such a degree, then the following would have happened,” where
one hasn’t lowered the temperature and the thing hasn’t happened, but one
has done something else and something else has happened. Many scientific
statements are made that way, and Goodman argues that there isn’t currently
a logic providing for them. But counterfactual conditionals are nonetheless
routinely used, and they are, nonetheless, enormously powerful. Which
suggests that perhaps a logic can be invented, or that they’re building on
something very strong.

Many uses of the private calendar are such uses. See, one of the problems
in developing a relationship is finding out that the states of the person you’re
with are to be accounted for by you , and not by the sheer fact that they’re with
somebody. That is, they want to be able to say that even if they were with
somebody before, they would still have been lonely. And that’s one wants to
do with these private calendars. They’re ways of building up, in deep and
repetitive ways, the relevance of ‘you.’ And perhaps one of the big things
about marriage is that that’s just what you’re constantly doing for each other.
The notion, for example, that marriage is made in heaven, is kind of an
underpinning to the use of these things. That is, it’s an account that would
provide the basis for saying “That was before I met you and I was lonely
then.” Our meeting was virtually guaranteed, and it’s just a matter of, until
then one drifted, and now it happened. By virtue of this causal structure, of
course, persons who are members of such units have built-in procedures for
finding that someone cares. And for a lot of things it’s the only way you can
find the sense of what’s going on.

Let me point out something about the private calendar that turns out to be
rather important. I don’t have a very large set of features of these things, but
one thing I have found out is that if we compare these private calendars to
everybody’s calendar, then there’s one striking difference between the two of
them. And that is, everybody’s calendar has, and private calendars do not
have, guaranteed continuity. Everybody’s calendar runs on into the indefinite
future, without regard to anybody in particular being present. Private
calendars end when ‘we’ end. The end of a relationship, in one way or
another, can provide that there’s no more events on the private calendar.

Now then, what we can see the widowed person saying, when they say

38

Part I

“Nothing happens anymore,’’ is that with regard to the private calendar
whereby events between me and my spouse happen and the value of my life
is found, no more events can occur on it. You can get, then, a sort of task that
a therapist, or somebody else to whom one of these persons would turn, might
have. The task is at least programatically simple, whether it’s easy to do is
another question. It involves bringing them back to the use of everybody’s
calendar, whereon events can still occur sensibly in their lives.

I’ll add another thing, and this is somewhat more conjectural, though not
strictly made up, and it may be relevant for our materials here. For widowed
persons, the fact that they’ve had a life with somebody is something that the
other’s death doesn't take away. And they can say “We had a marvelous 25
years together,” pointing to all the things we did together, how it is that I was
happy on this day because of what he was doing, because we were together,
etc. Now, when persons get divorced, something quite different seems to
operate. Apparently a divorce can provide for the fact that one can’t even
retrospectively use the private calendar one had going. The fact of a divorce,
perhaps with the reconsidering of whether one ever did care, and what after
all they were doing these last five years that led up to this, seems to involve
that one can’t then use it for the past that one was ‘together.’ That the woman
in our materials is divorced may then not only provide that she has no current
access to the built-in procedures for finding that others care which such a unit
as husband— wife provides, but also that she is deprived of whatever
retrospective use she might have had of that unit’s private calendar.

Via this sort of a sketch we can begin to see where the relevance comes of
having others care. And that is that the whole class of causal statements that
are built out of such units as husband-wife and the relationships between
categories in these units, provide an apparatus in which everybody is supposed
to be entitled to become a member of such a unit and thereby to have these
things done for them. And if they don’t become a member, given that they’re
entitled to become a member, they have a clear way of seeing that something
is missing. It’s not the easiest thing in the world to find a way to say that
something is missing. But if you have some objects for which there is no rule
of exclusion in the first place - everybody is entitled to them - then if someone
doesn’t have it at some point that one is entitled properly to have it, one can
say that it, and its consequences, are missing.

We can tie this up to some extent by asking what, then, is the consequence
of not having persons care? Well, these lay theories — and all these causal
statements and entitlement propositions are lay theories - have a rather
interesting property. If you consider our prototype of a scientific theory, then,
if some object doesn’t conform to what the theory proposes about the object,
then the theory has to be revised. This world has been constructed in a rather
more exquisite way. What goes on is the following. A large class of lay theories
are properly called ‘programmatic theories. ’ If they don’t describe your circum-
stances then it’s up to you to change. And if they don’t provide for you as a
Member, then it’s up to you to rid yourself of being a Member, for example
to kill yourself. In that way you keep the theories going as descriptive.

Lecture 5

39

If you’re a member of one of these units you have essentially automatic
ways of finding that others care. It’s built into the structure of ordinary
discourse, and the way persons see how events come off. If you’re not a
member of such a unit, it’s still relevant, but its structure is not available to
you. And you may then try that procedure which works for everybody -
dying - either as a way to find that somebody does care, or as a way of
providing that the theory that people ought to care is made correct by virtue
of your no longer being a Member. And we’ll see constantly that persons talk
of a whole range of things where if something is not so for them, then that
doesn’t provide that what’s supposed to be so is thereby wrong, but that
they’re wrong.

Let me add one more device relevant to “Does anybody care?” It is, of all
things, trash mail. The next time that they have hearings about removing
trash mail, I’m prepared to go and testify against its removal. Because trash
mail is a most interesting thing. I’ve mentioned this woman who used to go
to the park and sit and talk to people. Many of those were old ladies. They
were all utterly isolated. They came to Los Angeles after their whole family
died, or they came with their husband and he died. They live in apartments
near a park and they spend their day in the park. But they regulate their lives
in most interesting kinds of ways.

Even though they have almost no money they, for example, never purchase
at supermarkets and never purchase more than a day’s food. Because if
they did, they’d have nothing to do the next day. And they routinely will get
up - you’ll be sitting in the park talking to them, the only person who’s
talked to them since God knows when - they nevertheless get up and say “It’s
11 o’clock, I have to go home and check the mail.” Now there’s nobody
who’s writing to them. What it is, is that there’s that trash mail coming, and
that’s something.

Consider their situation: The mailman comes every day, and they know it.
And that means that for them, they have to go check the mail every day. The
only mail they do get is this mail that everybody gets. But for them, it’s
something. And if they had to recognize that he would come every day, and
every day they would find no mail, and they could look forward to that day
after day, then that situation of theirs, of isolation, would so be built into their
circumstances and shown to them routinely, that it might become far more
unbearable than it is - and it’s pretty unbearable - because this is a device that
happens every day, for whomsoever. You don’t know who is getting
telephone calls, you don’t know how many phone calls are being made, but
every day, everybody has the mailman go by. And if you just consider the
comparative cost of trash mail versus an enormous mental health operation,
then trash mail is not expensive. And for these people it’s by and large the
only means by which the routinely-used device of delivering mail does not
become the kind of thing it would otherwise become - this persistent
statement to them that nobody cares.

Lecture 6

The MIR Membership Categorization

Device

I’ll begin now talking about some very central machinery of social organiza-
tion. Let me indicate how I came by the findings I’m going to present. In
dealing with first conversations I’ve very frequently found, as anyone can
easily find, that especially in the early parts of these conversations certain
questions are prominent; questions like “What do you do?” “Where are you
from?” etc. I wanted to see if there was some simple way that I could describe
the items that those questions contain, so as to provide for their occurrence by
rather abstract descriptions. That was the initial task. Its consequences are
rather powerful, and I’ll develop them as I go along.

It seems that there is a class of category sets. By ‘category sets’ I means just
that: A set which is made up of a group of categories. There are more than
one set, each of which can be named, and they have common properties. And
that is what I mean by referring to them as a ‘class.’

A first thing we can say about this class of category sets is that its sets are
‘which’-type sets. By that I mean that whatever number of categories a set
contains, and without regard to the addition or subtraction of categories for
that set, each set’s categories classify a population. Now, I haven’t made up
these categories, they’re Members’ categories. The names of the sets would be
things like sex, age, race, religion, perhaps occupation. And in each set are
categories which can classify any member of the population. I call them
‘which’-type sets because questions about any one of these can be formulated
as, “Which, for some set, are you?,” and “None” is not a presumptive
member of any of the categories. And that would suggest what it is that
provides for such questions occurring at the early part of first conversations:
You don’t have to know anything about somebody to be able to formulate a
set of questions for which “None” is not an expectable answer. And of course
for some of the sets you don’t have to ask the question.

A second thing we can say about this class of category sets is that its
categories are what we can call ‘inference rich.’ By that I mean, a great deal
of the knowledge that members of a society have about the society is stored
in terms of these categories. And by ‘stored in terms of I mean that much
knowledge has some category term from this class as its subject. And the

A combination of Fall 1964, Ml, sides 1 and 2, and Winter 1965, lecture (4-5)
(transcriber unknown). The discussions are the same in both sets but are more formal
in the latter, and much of what is shown here is taken from the more formal version.

40

Lecture 6

41

inference-rich character of these categories constitutes another warrant for
their occurrence in early parts of first conversations: When you get some
category as an answer to a ‘which’-type question, you can feel that you know
a great deal about the person, and can readily formulate topics of conversation
based on the knowledge stored in terms of that category.

A third feature is that any member of any category is presumptively a
representative of that category for the purpose of use of whatever knowledge
is stored by reference to that category. So, for example, a foreigner comes to
the United States and you find yourself asking them about the political
situation in Ghana, or how they like the food in the United States, without
reference to whether they stand as a member of the Gourmet Club of France,
or don’t ever eat out, or aren’t interested much in food, or are just ordinary
citizens, so to speak. But one finds that it’s done. And it’s done for any of
these category sets.

Let me emphasize that we’re dealing with categories, and not necessarily
with what sociologists might call ‘groups,’ ‘organized groups,’ ‘organiza-
tions.’ It’s quite important to see that presumptive representativeness holds
whether or not the members of that category are or are not organized. If they
are organized, it holds whether or not they choose their representatives. The
fact that they are organized and choose their representatives does not mean
that one cannot apply the knowledge stored about such a category to persons
who are not selected by the group. And, furthermore, the fact that it is not
a group in the sense of being organized, doesn’t mean that one cannot apply
such knowledge.

I’m calling this whole apparatus the MIR device. And that is an acronym.
‘M’ stands for membership. T stands for inference-rich, and ‘R’ stands for
representative. That’s the core of the machinery. I take it one can readily
notice how absolutely central this is, for a vast amount of stuff is handled by
Members in terms of the categories that it locates and the way it locates them,
and the activities that those categories are used to handle.

Now one might get a sense here of a certain problem, and I’ll mention it
right now. I take it to be a central problem of sociology, and I’ll try to show
some sorts of solutions to it eventually. The problem is this: There are these
category sets. For any person being talked of, how is it that Members go about
selecting the set in terms of whose categories that person is going to be talked
of? It’s perfectly obvious that there is a range of sets whose categories could
be used; from the set ‘sex,’ “a woman”. From ‘race,’ ‘‘a Negro.” From
‘religion,’ ‘‘a Catholic.” From ‘occupation,’ “a psychiatric social worker,”
etc., etc. Each of these categories could apply to the same person. And it’s
perfectly obvious that Members do use one set’s categories for some
statements and another set’s categories for other statements. If we’re going to
describe Members’ activities, and the way they produce activities and see
activities and organize their knowledge about them, then we’re going to have
to find out how they go about choosing among the available sets of categories
for grasping some event.

All the sociology we read is unanalytic, in the sense that they simply put

42

Part I

some category in. They may make sense to us in doing that, but they’re doing
it simply as another Member. They haven’t described the phenomena they’re
seeking to describe - or that they ought to be seeking to describe. What they
need to do is give us some procedure for choosing that category which is used
to present some piece of information. And that brings us back to the question,
are there procedures that Members have for selecting categories? One of my
aims is to show that there are.

For now, let me show one of the tasks this MIR device, in combination
with a particular sort of operation, can be involved in. I’ll be talking now
about some extremely basic and extremely generic social control devices. The
particular sort of operation consists of one way that Members go about
making new knowledge. Suppose some event occurs and is known about by
reference to the name of the person who did it. The way you get a piece of
knowledge involves pulling out the name and putting in some category. Then
one gets, not ‘John did X,’ but ‘a such-and-such did X.’ In that way one gets
additions to any given body of knowledge about such categories. And what
we find is that an enormous amount of what we could call the lay theories of
social actions are fitted onto these categories.

Given the MIR device, and given this operation whereby new knowledge
is formulated - by replacing a name with a category - we can begin to see
how a class of social control devices gets set up and is used. It has as its basis
that members of the society are constantly engaged in monitoring events; on
the one hand by reference to whether something that has happened is
something that they’re accountable for, and on the other hand, to find out
what is getting done by members of any of the other categories.

Apart from the routine monitoring terms of these categories, we get nicely
special kinds of occurrences which provide a beautiful view of tensions arising
as persons await the discovery of which of them is going to be found to have
done this thing. For example, the hours between the assassination of President
Kennedy and the determination of who it was, and thus what category it was
that performed the act. If you have access to a variety of materials from that
time, you can see persons reporting themselves going through “Was it one of
us right-wing Republicians?,’’ “Was it one of us Negroes?,” “Was it a Jew?.”
etc. That is, “Was it me?” in that sense.

This sort of monitoring makes for great sophistication in kinds of ways
of doing trouble. For example, in the recent Russian economic trials, it was
quite sufficient for those who were encouraging anti-semitism in Russia, to
simply publish the names of the persons who were tried. The names turned
out to be seeable as belonging to Jews. And you could leave the rest to
everybody’s routine procedures: “See? Jews are economic criminals, as
everybody knew.”

What you get then is a whole series of teachings, all of which have the same
form: “Remember you’re a such-and-such” (a lady, an American, a Negro,
a Catholic, etc.). That is, any action you take is exemplary. Any action you
take is something we’re going to have to come to terms with. Such teachings
belong to a class of activities which are often called ‘internal systems of social

Lecture 6

43

control.’ They may be informal or formal, but they’re not like such things as,
for example, a government. What they have in common is that they are all
operated by and enforced by and taught by and used by members of the
category whose members are to be controlled.

These internal control devices all seem to be built on, and have their power
by virtue of, this very simple apparatus the MIR device, which is utterly
disjunctive to whatever these groups happen to be, or whether they happen to
be ‘groups’ in the organizational sense of the term. Indeed, some of these
classes’ categories may set out to become organizations by virtue of the uses
that are being made in any case of their categorial membership. It is no mere
plea, for example, that women have got to get organized. If an event occurs
where what somebody does is seen by reference to the category ‘female,’ then
women in the society find themselves constructing explanations for how it
would have happened, or proposing that they can’t figure out how such a one
would have done that - though of course they need know nothing about the
person except that category of membership.

To get some contrast on these control devices, we might notice that they’re
much unlike a characteristic thing to which we give the name ‘scapegoating.’
In scapegoating, if a member of a group does something which is sanction-
able, then either that member or some other member of the group may be
selected by some procedure - perhaps the group itself selects them, perhaps
an outsider selects them - and a sanction is applied to that person or persons.
That being done, guilt is purged. This doesn’t happen. If somebody does
something that is formulated as an action of the group, then a piece of
knowledge is thereby legitimately usable and you don’t have some procedure
for cutting off its use - such as the scapegoating procedure - where it’s
thereafter not proper to say about a group that they did this thing. So,
Members’ stake in the actions of other Members is not conditional in that
sense; that is, it is not conditional on being or not being purged.

Now, given what I’ve said so far, there are certain things that can be
examined pretty clearly. One of them has to do with a situation where some
person who is a member of some category does something to another person,
where that other person happens to be a member of the same category. The
pain of it, that it’s something awful, operates not merely by virtue of the fact
that somebody did something to you for which you have some legitimate
feeling of being injured, but in addition you have the sense of having injured
yourself. And in such cases people talk of being ashamed of being a member
of the category they’re a member of.

So, for example, I have materials in which a person in one or another
trouble will call the social agency of their religious denomination (Catholic
Family Service, Jewish Welfare Group, etc.), and get rejected. I have two
cases in which such a person subsequently calls someone else and is reporting
this. And in each case they report - as is not reported when persons are
rejected by, say, a municipal agency or some hospital - “It’s not just that I
was rejected by somebody, but I feel ashamed of belonging to a group that
does that to people.’’ That is, impersonally they are now observing the group

44

Part I

in operation, where they take it that whoever it was that spoke to them is a
representative of the category (Catholic, Jew, etc.), and that such a one did
something to another is something that reflects on themselves and devalues
their membership. That is, not only non-members, but members of a category
take it that the actions of that category can be assessed. It’s not merely that
a non-Catholic could hold this up and say, “See? Catholics don’t take care of
their own,” but that a Catholic will say, about their own group, the same
thing. The generic importance of such a phenomenon is that it’s not just one
category’s view of another, but that knowledge is standardized across the
categories.

I want now to notice several affiliated features of the MIR device. I’ll start
off with a feature I found in a few pieces of conversation - and indeed it was
looking at these materials that really launched a lot of the considerations here.
First I’ll show the materials. In the first extract, two teenagers, a fellow and
a girl, are talking about dating.

A: Corliss the g- this chick that I’m hanging around with now, she’s real

nice she’s got a real good personality, she’s not- you know she’s //
just a real cute little kid.

B: Mm hm,

A: And last night we went to the Mardi Gras together and we were both

well we were both pooped because I-I ran in the track meet yesterday.
And she-she’s in the girls’ tumbling team. I mean she doesn’t like it
she’s just on it because she needs the credits.

The second is from the emergency psychiatric hospital.

A: How old are you Mr Bergstein?

B: I’m 48, I look much younger. I look about 35, and I’m quite

ambitious and quite idealistic and very inventive and conscientious
and responsible.

What struck me was that a thing was presented, “She’s on the girls’ tumbling
team” and “I’m 48,” and then you get a modifier in a long or short
statement. At this point in our considerations I take it that what we have here
is very apparent. For some category, like these two, there is a set of things
known about a member which can be applied to any member, for example
that being on the girls’ tumbling team is presumably something very gauche,
and that someone who is 48 is past their prime, which any person now talking
about such a category membership has to come to terms with.

So we have a class of things, these modifiers. And they consist of attempts
to provide that what it is that may be said about any member is not to be said
about the member at hand. “I’m 48 but I look and feel younger,” “She’s on
the girl’s tumbling team but . . “He’s a Negro but . . . ,” etc. Having
seen this, I think we can see a solution to something I had been puzzled about
for quite awhile. It is, I suppose, a rather minor kind of interest, but it’s

Lecture 6

45

related to just this business. There is a class of what look like tautological
statements, “Women will be women,’’ “White folks is white folks,” a whole
large bunch of them. It looks like they are simply tautological statements; that
is, that they say nothing. But given the use of MIR modifiers, these
tautological statements seem to be - to use an awkward phrase - ‘anti-
modifier modifiers.’ That is to say, under a condition where for some reason
it’s proposed, or one has been going along under the notion that, the person
whose behavior is being considered is to be classified by reference to one of
these modifiers - for example, “He’s a Negro, but the things you can say
about Negroes you can’t say about him” - you have this other class of
statements available to flip in and provide that in the last analysis he’s like the
others. They provide for the re-relevance of whatever it is that’s known about
the category. And if you watch conversations in which these things occur,
that’s the way they get used.

In certain kinds of relationships these anti-modifier modifiers can be
extremely deadly kinds of things to have around, because they can always be
used in the following way. There is the kind of relationship that proceeds
between some set of persons under the modification that what is known about
the category that one of them is a member of, is not to be applied to this one.
And now the one who is living under that modification has always to carry the
possibility that some time somebody is going to say, “Well, now it really
comes out ...” and invoke a set of things that can properly be said about that
category, removing the modification under which they’ve been living.

There is another feature I want to mention. Some of the category sets of the
MIR device have to be differentiated from others in some special ways. If one
considers categories like age and social class, in contrast to those like race and
sex, one finds some rather interesting differences. For all of them there may
well be a stable set of categories used by everybody. But whereas for, say, sex
and race it will be by and large the case that one can take it that whatever
category somebody applies to somebody else or to themselves, anybody else
would apply that category, that is not so for categories like age and social
class. What you have with these latter sorts of categories is a rather lovely
series of things going on. If any Member hears another categorize someone
else or themselves on one of these items, then the way the Member hearing
this decides what category is appropriate, is by themselves categorizing the
categorizer according to the same set of categories. So, if you hear B categorize
C as ‘old,’ then you would categorize B to decide how you would categorize
C. And again, the same procedure works for such a thing as social class.

This sort of operation is probably basic to something which sociologists are
talking about as a generic matter (but which is by no means generic) and which
Members also use. And that’s the notion of ‘perspective.’ If somebody calls
somebody ‘old,’ what you want to know is, how would you call the person who
called the person old? That way, even if you don’t know the person being
categorized, you can have, for example, some notion as to the range that’s
involved. If you’re an adult and it’s a ten-year-old who calls somebody old,
then you can figure that the somebody could be anywhere from 20 on up;

46

Part I

that is, they’re possibly young. Now it may well be, though I can’t say this
with any confidence, that even though this sort of business is going on, it’s
largely the case that the same information is stored for any category. So that
A and B, being from different categories, may place C in different categories
(A categorizing C as old, B categorizing C as young). But whoever A and B
would place in the same category (whoever A might call old and whoever B
might call old), they would say the same things about that one. So that when
kids talk about ‘old people,’ though they’re talking about somebody who is
20, they may have the same information about that one that a 30-year-old
has about someone who is 70.

It may not be the case, then, that on the different ‘perspectives’ - that is,
on the different uses of the categories - depends a whole different body of
information, but that the knowledge is stable for any category like ‘old.’ It’s
just a matter of what category is using the term ‘old.’ The same thing may
work for social class. For these, there may be no position that provides for the
definite classification of somebody. One wouldn’t then find that somebody
carries around an identity which is stable for any environment they come into;
for example, that they would be ‘old’ no matter who it is that’s around them.
Nor would it be the case that for each of the persons around them they would
be seen as the same person that they have to see themselves as. There is no
supposition of agreement on any categorization for such persons all catego-
rizing each other.

Let me mention another minor thing which this machinery can clarify.
Where it might not be proper to say certain things about another person, or
for that matter about oneself, what one can do is to propose membership in
some category, where that category stands as the adequate basis for inferring
those certain other things. So, to take a sort of extreme example, here is a first
conversation between a psychiatric social worker and a suicidal man. Earlier
the man had said that he’s been married three times, he’s not working and not
married now.

A: Is there anything you can stay interested in?

B: No, not really.

A: What interests did you have before?

B: I was a hair stylist at one time, I did some fashions now and then,

things like that.

They go on for a couple of minutes. Then:

A: Have you been having some sexual problems?

B: All my life.

A: Uh huh. Yeah.

B: Naturally. You probably suspect, as far as the hair stylist and, uh,

either one way or the other, they’re straight or homosexual, something
like that.

Lecture 6

47

In this case, while it might not be proper for this man to say about himself
that he’s troubled by possible homosexual tendencies, he finds a way to
invoke a subset of occupational categories, “hair stylist. . .fashions. . .and
things like that,’’ which constitutes an adequate basis for inferring homosex-
uality. And in his subsequent talk he proposes that such an inference has
“probably’’ been made by the other. Apparently, then, there are ways of
introducing a piece of information and testing out whether it will be
acceptable, which don’t involve saying it.

Now, I have by and large been talking about negative information stored
in these categories, but they obviously provide that system of rewards which
any young person can expect by virtue of becoming a member of any category
that they can become a member of — for example, occupation, or changes of
religion - since to become a member is to make stateable about yourself any
of the things that are stateable about a member. So they’re the basic system
of incentive for persons to do a variety of things. That’s now fairly well
known, so all the various occupations are engaged in trying to sharpen up
their images so as to make it attractive for persons to come into them.

Also, there are paths that can be constructed in terms of set movement,
which persons may use to consider what their likelihoods are of ever getting
to be a such-and-such, given that I’m a this-and-that. And the fact that
‘somebody has made it’ provides a path. But this can be used in very negative
ways just as well. For example, I heard on the radio the other night an
interview with congressman Billy Mills. And he was asked, “Well, you’re a
Negro, you came from the lowest part of the economic structure and
nonetheless you’ve made a great success, why can’t any other Negro do it?’ ’
That is, there is a path which your history has laid out. Why does that not
stand as a way to get from one place to another which any person in your
initial state can use? And now Mills has to come up with some argument
about why it couldn’t or shouldn’t. Or, for example, the notion of ‘token
integration’ can be used in business establishments in just that way. The fact
that there is a such-and-such at this place provides that there is indeed a path
from the position they started with. And since there is a path, and since
Members at that initial position know how to use it, the burden now lies on
them to do so.

Finally, let me offer something to consider. I have no idea whether it’s so.
It sounds altogether too smooth to me, and nonetheless it also looks, on the
face of it, to be very descriptive. Many of these classes are, or can be built as,
two-set classes. Sex is a two-set class. Race can be formulated as a two-set
class; for example, non-whites and whites. And there’s rich and poor, old and
young, etc. The question I’m asking is, does it matter, for the kinds of things
that can be done with these classes, how many sets they contain? Two-set
classes would seem to have certain kinds of attractions. For example, they’re
tremendously easy to compare. With a two-set class you can apparently make
an observation of comparative lack much more easily than otherwise. And I
wonder, for example, whether many kinds of conflict and perhaps most sorts
of revolutions occur by virtue of these two-set classes; as we say, the haves and

48

Part I

the don’t haves. Under such a view, you can see all sorts of different things
being fitted to the notion of haves and don’t haves. Marx can be seen to have
used this two-set class. The movement for equality of women can be seen to
be using it. And the Negro revolution as well.

To establish a two-set class you might start with one group who you locate
as the group in power, or the haves. Give them a name: Whites, men, the old.
And then assimilate all the others to some predominant feature of those
others; for example, a lot of them are Negro so you call it ‘the black
revolution.’ But if you just go through the ways revolutions tend to organize
themselves, or the ways movements tend to organize themselves, or notice
that games - which are model conflict situations - are so often either two-
party systems or variants of two-party systems, it begins to look as though
formulating in terms of two-class sets is a method of doing things. Whether
this is so or not, I haven’t the vaguest idea and it needn’t be taken with any
seriousness. But what I said earlier, about this device as a basic mechanism of
social control is, I think, important. If we can come to see what’s involved in
it, I think we can see something useful, and something that’s of theoretical
importance in sociology.

Lecture 7
On Questions

A few days ago I asked people to go out and collect first lines of ‘pickups.’ I
did that because I want to return to some considerations of rules of
conversational sequencing, and to introduce a rule. But it has initially such a
vacuous sound to it that I wanted to see if there was some way of making its
relevance clear, more quickly and more intimately than the materials I had on
hand permitted. Looking at my materials, these long collections of talk, and
trying to get an abstract rule that would generate, not the particular things
that are said, but let’s say the sequences, or their continuity and things like
that, then you come up with a rule that says something like: A person who asks
a question has a right to talk again , after the other talks.

That sounds enormously empty. One reason it sounds empty is, at this
point I can’t put in “. . . after the other answers.” I can’t put it in, in part
because ‘questions’ and ‘answers’ are in some ways altogether different
objects. For example, a question is a grammatical matter and an answer isn’t.
I don’t think you can locate ‘an answer’ grammatically. Also, a question can
be paralinguistically described. An answer doesn’t have that character. That
is, if you play a tape or listen to a conversation and forget about what’s being
said but just try to get the tone, pitch and that sort of thing, then you can
describe what a question looks like. A question has a form. And an answer
doesn’t, apparently. So we can talk about ‘asking questions’ and identify some
object as ‘a question,’ but we can’t do that very much with ‘an answer.’

But anyway, there looks to be a rule that a person who asks a question has
a right to talk again afterwards. And that rule can provide a simple way of
generating enormous masses of sequences of talk: Question, talk, question,
talk, etc., etc. We can say it’s a rule with a repeat device. But what else can
we get from it? About 60 first lines of ‘pickups’ were handed in. Seven of
them were other than questions. More than 50 were questions. And by
reference to this rule, that wouldn’t be an incidental fact. You might begin to
see why questions get used.

Here’s a classic Yiddish joke on just this issue. A young man gets on a train
and sits down next to an older man. The younger one asks, “Can you tell me
the time?” and the older man says “No.” “What do you mean no?” the
younger one says. The older one says, “If I tell you the time we will have to
get into a conversation. You’ll ask me where I’m going. I’ll ask you where
you’re going. It will turn out we’re going to the same place. I’ll have to invite

Fall 1964, M2, side 2 and ‘(Fall 1964)’, pp. 1-9 with a brief extract from
Fall 1964, tape 7, side 1, p. 3.

49

50

Part I

you for dinner. I have a young marriageable daughter, and I don’t want my
daughter to marry someone who doesn’t wear a watch.”

I’ll make something of an excursion and talk a little bit about pickups. Let
me sort of mock one through; one that I saw at the airport. A bunch of about
20 people are standing around waiting for a plane to arrive. At the edge of
the crowd a girl is standing. A guy comes up somewhat behind her and says,
‘‘What time does the plane arrive?” She turns and says ‘‘In 20 minutes” and
turns back. Then he asks another question. She turns, answers, turns back.
This goes on for five or six questions. Then she just turns her body to him,
without especially any expression, or even looking at him. He keeps asking
questions. She keeps answering them. She turns her head up when she
answers and then brings it back. So it looks like if at any point he was to stop,
that would be that. At some point she takes out a cigarette and he lights it.

Now, it’s a very well known fact about homosexual society that in bars
frequented by male homosexuals the bartender keeps a pack of matches
behind the bar because it’s impossible to ask for a light. That is to say, asking
for a light is so much a ‘move’ that if what you want is just a light, you just
pretty much can’t do it. It also seems that between males and females, asking
for a light is a tremendously sexual thing to do. In any case, in such things as
pickups it’s a key point. And I don’t think that’s incidental. There are a series
of areas which are closely regulated, especially between persons who don’t
know each other, which lighting a cigarette can be involved in.

First is eye-to-eye contact. If, for example, you sit in the library and have
somebody sit opposite you watching you, you can get the following results.
You start to look at somebody. You’ll find that even when you can’t tell that
the other person has caught your glance, the person sitting opposite you can
- and they’re not looking at the other, they’re just looking at you. Your eyes
flick, like magic, across the room. A glance just drives them away. So
eye-to-eye contact is a highly regulated thing. But of course when a person
lights a cigarette for you, that’s an occasion for persons to align with heads
very close to each other, looking at each other directly.

A second thing is touching; a very closely regulated matter as between
persons who don’t know each other, and even between persons who know
each other. But again, lighting a cigarette can be an occasion for taking
someone’s hand, for example. In any case, some touching can very easily go on.

It’s also a point when business is being done between the persons such that
they’re not talking and nonetheless together. Whereas previously, every time
they stop talking there’s an issue about whether they are together. But here
you have a first occasion when the persons can stand together, looking at each
other perhaps, without ever saying anything, and pick it up again. And the
development of the ability to have a pause in a relationship is very crucial.
Just remember, for example, when we were very young, going out on dates
for the first times, how it was that gaps in conversation were treated as such
tremendously painful things. The developing of an ability in a relationship to
be silent is a very important thing. And this is a first step to that.

Another thing this act of getting a cigarette lighted does - this is of course

Lecture 7

51

completely apart from anything like psychoanalytic theories about cigarettes

- is that when a woman has a cigarette, she has an occasion to be moving.
You don’t smoke a cigarette with your hands at your sides. And, for example,
as the cigarette is lighted for her, she can bend her head and look up through
the top of her eyes - which is in our society a very sexual kind of way to be
looking at somebody. And when she has a cigarette in her hand, she has then
a routine opportunity for bringing her hand to her face, which provides the
minimal condition for him to be looking at her face; that is, it’s perfectly
legitimate to follow a moving object, your eyes do it normally. So she can
begin to get into physical action, moving her head and her arms, etc., which
then provides for, literally, a dance.

That ends the excursion. I wanted to give some idea about the relevance of
this rule that if you ask a question you have a chance to talk again afterwards.

Now, there are some situations which have their particular character by
way of the fact that this rule doesn’t hold. The easiest one to get examples of
are Presidential press conferences. A bunch of reporters are in, say, the State
Department auditorium, and one of them is pointed to. They can ask one
question. Then somebody else will be pointed to. This isn’t so for Johnson’s
recent conferences because he’s been doing them informally; people wander
around together. And there it seems to be permissable for a fellow to ask a
second question, a clarifying question. But in general, you could take the
Presidential press conference as one extreme: You only get one question. One
thing you could then do is move to the other extreme: Places where one has
an enforced right to ask further questions, a right which is not violatable by
the person who is answering - cross examination, for example. Then you
could try to see what questions look like under this or that control. For
example, do they have a different shape? And what’s the position of either the
questioner or the answerer in each situation?

Here’s one of the problems for a questioner at a press conference. Persons
come in with questions. They all move to be called by doing various things

- standing up, waving their arms, getting front seats, wearing odd clothes,
etc. One of them gets recognized. He asks a question. An answer comes out.
If he had an opportunity to talk again, he might want to push further on what
that answer suggests. But he can’t. Somebody else is now called on. Now their
problem is, they may know that there’s a good question one could make up
to follow that one. But if they follow that one, what happens to their
question? They may have come with something of particular interest to their
area, “What’s your policy on Idaho potatoes?’’ They want to get that out. But
what’s just been raised is something about the defense policy, which is a
generally important issue. Should they ask their question or the follow-up
question? They won’t get another opportunity. Now for some questions
they’ve brought with them, they can know that these are so topical that
another might well raise it, and they can ‘sacrifice’ their use of it to doing
what’s in the interest of the press conference; that is, asking a follow-up
question. But of course the President has some control over what questions get
asked in the first place, because he will have some pretty good idea about

52

Part I

what various persons might be interested in asking. (Kennedy apparently
sometimes gave them the questions, or they told him what they were going
to ask if they were recognized.)

Another sort of problem for a questioner at a press conference is that the
answer might be nothing more than a “Yes” or “No.” Or, for example, the
Kennedy format, responding to a question with a quip. And that in its way
is akin to “This is Mr Smith may I help you?” as a way of getting the other’s
name without saying “What is your name?” and also akin to laughing when
someone says “I’m going to kill myself’ as a way of refusing to give help
without saying “I won’t help you.” The Kennedy quip is a way of refusing
to answer a question without saying “I’m not going to answer that question.”
Because after the quip, the fellow who asked the question can’t come back.
And whether anybody else will come back to it remains to be seen.

What might stand as a solution to this problem was something that was
noticed when they were thinking of televising the press conferences. One of
the matters that arose was, if they televised these things, people would ask
very long, involved questions to get themselves time before the audience.
Then it was realized, well they do that anyway. Now the question is, what are
these long involved questions doing? One thing they might be doing is this:
You try to ask a question, providing in the question for a sub-answer:
“Would you do so-and-so or not, and if not, then . . . ?” That is, you’d want
to provide that he could not just say “Yes” and you wouldn’t have anything,
but “If ‘yes,’ then is such-and-such relevant, and if not, is such-and-such
relevant?” These sorts of questions sometimes get so involved that after one
of them the President says “Well you’ve made your point, thank you.”
Because it looks like a speech. Now that elaborateness may not be so much
due to the guy’s being on television as to his having to make a question under
this constraint of not having a second chance, to ask for a clarification of the
answer or to pose a follow-up question. You have to try, then, to build the one
question to do that job, whereas in other situations you can just ask them piece
by piece.

So we can begin to see how things might look different by virtue of the
presence or absence of the rule and how strongly that is enforced. For
example, in cross-examination the presence of the rule is strictly enforced.
You’re sworn in and you stay on the stand until the guys says “No more
questions”. You can’t say “Well that’s enough, don’t bother me anymore,”
which you can do in ordinary conversation. And, for example, in ordinary
conversation you can produce an answer, append to it a question, and now
take over. The roles are not set so that one is ‘a questioner’ and one is ‘an
answerer’ from now until the end of time, or from now until the end of this
interchange. So something as limited as this rule can have rather considerable
and observable consequences.

I’m going to introduce another rule. I’ll start out with some quotations. 1

1 Cases (l)-(4) are from Sacks’ research notes. He often put big blocks of data on
the blackboard, usually reading it out. He does not do so this time, but he refers

Lecture 7

53

(1) A: Do you have any physical problem? Any illnesses?

B : No. Just a little bit ( ) and overweight.

A : Overweight?

B : Just about 20 pounds.

A: Now when you get depressed, how is your appetite?

B : It’s usually always good. No when I’m depressed I don’t feel like

eating, I can’t eat.

A: And how do you sleep?

B : Alright, but not as much as I’d like to.

A: That happens. Do you lose much interest in sex when you’re

depressed?

B : Well, just in everything when I’m depressed.

A: Now, right now you’re feeling a little better.

B : Yes.

A: But you figure on past experience that you might become

depressed again any time, like tomorrow.

B : Yes.

A ► Well, it certainly sounds as if you need some help, and there are
many things that can be done for this.

(2) A: Are you working?

B : No I’m not.

A: Your husband supports you?

B : Yes.

A: Well, what do you do with yourself?

B : Oh, I have a lot of interests. I work with theater. I do, oh, little

community theater direction and things of this order.

A: Do you find there are times when you lose interest in it?

B : Yes. Very d- I find there are times when I lose interest in

everything and there have been times when I have stopped
speaking for days.

A : I don’t know anything about your sex life now, but are there times

when you lose interest in sex?

B : Yes. Completely so.

A: - *Right. Sounds pretty clear cut as a depressive illness.

(3) A: Hello, this is Mr Smith.

B : How does the Emergency Psychiatric Hospital work out?

A: How does it work out?

B : I mean what do you do. I’m sure a telephone conversation

wouldn’t save me.

A: Well, it sometimes is a first step in the process, going a very long

way in being helpful.

to the materials in such a way that at least two of them can be located with some
certainty, and the other two stand as decent instances of what is being talked of in the
lecture.

54

Part I

B : A first what now?

A: A first step.

B : In a telephone conversation. Then what’s the next step?

A : Well, we would If it’s a problem involving suicide we would invite

you to come in for an interview, explore the problem more fully
with you, and then see if we could recommend something helpful.

B *-It sounds too slow.

A: What’s that I’m having difficulty hearing.

B : I said it sounds too slow.

A: Too slow, we can act very rapidly when the need arises. Are you

a person with a suicidal problem?

B : Yes.

A: Could you tell me something about it?

What we find in these exchanges is that the person who is asking the questions
seems to have first rights to perform an operation on the set of answers. You
can call it ‘draw a conclusion.’ Socrates used the phrase ‘add them up.’ It was
very basic to his way of doing dialectic. He would go along and then say at
some point, “Well, let’s see where we are. Let’s add up the answers and draw
some conclusion.”

And it’s that right that provides for a lot of what look like strugglings in
some conversations, where the attempt to move into the position of
‘questioner’ seems to be quite a thing that persons try to do. (We can just note
that in our third exchange there’s a shift in the middle, A, the answerer,
appending to his answer a question, which B answers, and then a next
question.) As long as one is in the position of doing the questions, then in part
they have control of the conversation.

Now, we do find questions followed by questions. For example, you can
propose a request for clarification of the question.

( 4 )

A: This is Mr Smith. May I help you?

B : Yes, I heard that you help people who are on the verge of

committing suicide or something.

A: What was that?

B : I said that you help people who are on the verge of committing

suicide.

A: Yes, we do, certainly.

B *Tn what way?

A: - ►Well, let me ask this. Are you calling about yourself?

B : Yes, uh huh.

A: Well, what we do is talk with them. We usually invite them to

come into the office, try to determine what the problem is, and see
in various ways how we can help by talking understanding the
problem and sometimes making recommendations sometimes
getting them in contact with a psychiatric person to help them.

B : I see. Where is the office?

Lecture 7

55

But as compared to ceremonials - where once the slot in which the next piece
of talk goes is filled with something else, then the ceremonial doesn’t hold the
floor, so to speak - I don’t think that’s so for questions. For example, here’s
one where, after two questions by someone who had been asked a question,
the first question is answered.

A: Hey did you talk Marcia into coming down here?

B : Was she here?

A: Yeah.

B: When did she leave.

A: About a half hour ago.

B: Yeah, I talked her into living here with me.

So an answerer can ask a question after a question. But what happens is, you
get a parenthesis in the conversation. There is a question on the floor, and an
interlude in which some question or set of questions operates, where, once that
is answered the initial question comes back into its relevance as controlling
what’s supposed to go next.

As I said, as long as one is doing the questions, then in part one has control
of the conversation. In our first four instances the person who is calling is
doing so, presumably, for help. One difference in them is that in the latter
two, the caller starts holding the floor with a series of questions. And by doing
that, they have the opportunity to assess whether the proposed help will be
something that can help them and whether they’re going to use it. They can
say, as we get in the third instance, “It sounds too slow.” In the first two, that
doesn’t take place. And what one has in the latter two is a situation where the
caller is in the driver’s seat; they can hold back asking whether they should
come in, and go about deciding that on the basis of the answers they get.
Whereas in the first two, the recipient of the call gets the materials, on the
basis of which they can propose that the other ought to come in for help.

There is a lot more detail about how, in particular ways, the bringing of an
agreement or the holding off of an agreement to come in, takes place. But the
outline of it, by reference to this rule that persons asking questions have a
right to make the first operation on the answers, is the first thing I wanted to
see. You can then watch for it happening in altogether different kinds of
scenes.

Let me now focus on the fact that the one who is doing the questions has
control of the conversation, in part. There can be a sense in which, while
you’re asking the questions, you could not be said to be in control. I’ll give one
of the best examples I’ve ever seen of this sort of thing. I took a course called
Constitutional Litigation with an enormously smart man, a guy named
Telford Taylor who was chief prosecutor at the Nuremberg Trials. The first
day, we came into the class and there were twelve people. He went through
the list of persons who were supposed to be there. Thirteen had signed up for
the course. The door opens and somebody walks in, just as Taylor has finished
going through the names. He says “Mr Jones.” The guy says “No.” Walks

56

Part I

over and sits down. Taylor says, “You’re not Mr Jones.” “No, I’m not.” So
he’s used the correction-invitation device, asking “If you’re not Mr Jones,
who are you?” without saying that. In return, he’s gotten “No, I’m not.” So
he says, “Well, on the list here, Mr Jones is the only one who isn’t present,
who is signed up.” The.fellow says, “I didn’t sign up.” Taylor says, “Well I
ought to have the names of all the people who are taking the course.” The
fellow says, “Well I’m not sure I’m going to take the course.” Taylor says,
“But even anybody who just happens to be here!” “Well, maybe I’ll leave.”

So then Telford Taylor sits and scratches his head, says “I’m sure if I go to
the Dean’s office I’ll find a rule that says if I ask somebody for their name,
they have to give me their name.” The fellow says, “You haven’t asked me
for my name.” At which point the class starts wildly laughing, and Taylor
points and says, “Now sometimes you get a witness like that!” And you do,
of course, get just this kind of thing; the ability to extend a series of returns
which avoid what’s being asked for but hasn’t been asked for point blank.

Now, in contrast to that, we get something we can call an ‘answer
constructed by reference to the project of the question.’ We’ve talked about
this correction-invitation device. What you got there was, for example, “Is
your husband a police officer?” to which you could answer “No” or “No, he’s
a such-and-such.” And it’s the latter answer which we can call an answer
constructed by reference to the project of the question. That is, what you can
see that the question wants to find out, is something that controls how you
answer it. Here is another instance of such an answer.

B: I just thought you sent somebody out to talk to somebody. You

know, Twelve Step type of thing. Because I must have a-

A: Have you ever been in AA?

B: I don’t drink.

A: You don’t drink?

B: I mean, I drink if I’m out at a party and everybody’s having a drink.

I take one to everybody’s two, but I’ve had friends . . .

That is, the caller has mentioned that this clinic might work like Alcoholics
Anonymous (Twelve Steps being one of their procedures). Asked if she’s ever
been in that organization, she gives this answer, “I don’t drink.” She could
have said just “No.” But she understood what it was that he wanted to find
out, and produced an answer by reference to that.

These sorts of answers take a kind of cooperation which is not present when
the answerer leaves the questioner ‘in control’ of the conversation - which can
also be characterized as letting the questioner go off on as many wrong tracks
as he pleases, where you can get a long, involved project that generates a series
of questions, none of which turn out to have any use.

Lecture 8
On Measuring

I’ll begin off with some quotations. 1

(1) A: People have to be pretty upset to want to kill themselves.

B: Oh I have been. Don’t think I haven’t.

A: Have you been sleeping?

B: No I haven’t.

A: And how’s your appetite.

B: Well, not too good.

(2) A: Let me hear a little more. How’s your appetite.

B: Not too good.

A: Yeah. Haven’t felt like eating?

B: ( ) other day

A: And you haven’t been sleeping too well?

B: Not very well.

(3) A: How are you sleeping?

B: Well, that’s a peculiar part, I sleep very well.

A: And how is your appetite.

B: My appetite is very good. That’s what I don’t understand.

A: You sound as though you know something about the symptoms of

depression.

I want to notice the following sort of thing. Recurrently in these
conversations you get a question like “How are you sleeping?” “Not too
good,” “How’s your appetite?” “Not good.” Now it looks like absolutely
nothing is going on. And quite frankly when I thought I was essentially
finished with analyzing this stuff, I hadn’t seen something that’s going on
there that is utterly fantastic. Absolutely unbelievable in a way. And let me
just try to begin to show what it is.

I’ll start out with something I’d already worked up, which just begins to
set this problem. And that is, there are a variety of items on which persons can

A combination of extracts from several lectures: Fall 1964, tape 14, side 2,
pp. 1-10, Ml, side 2, pp. 16-17, M2, side 1, pp. 1-4, 7-12, and 14-18, and
Winter 1965, lecture (4-5), 14-15 (transcriber unknown).

1 Only the first of these extracts is on the tape; the other two are taken from Sacks’
research notes.

57

58

Part I

monitor their own states. They either do it routinely, or, when asked, they
have the information which permits them to engage in some consideration
and then give an answer as to a current state or current variation from a prior
state. The items are things like sleeping, appetite, etc. And all I’m saying,
then, is that persons can go about monitoring how they’re eating and sleeping,
on those questions, they know.

Now for each of these there is a category that Members use, and that
category is ‘normal.’ That is an extraordinarily special category. Each Member
employs it. Questions can be asked about it from one to another. But for a
large variety of uses of the answer to questions within which ‘normal’ might
be an answer, and for a large variety of the monitorings which might go on
by oneself without anybody asking you anything about it, it’s quite irrelevant
what, for any given person, that notion ‘normal’ denotes. It’s irrelevant
whether it’s similar to or different from the features that anybody else uses to
decide that on some item they’re ‘normal.’ So ‘normal’ is a standardized
category, where whatever it refers to for any given person doesn’t have to be
specified to control its use.

And there are variant categories which are also standardized, like ‘poor’
and ‘great.’ There are whole bunches of terms: Fantastic, cool, terrible,
whatever you want. It’s essentially the same. We’ll call them ‘directional
differences.’ Minus and plus. So we get ‘normal,’ ‘minus’ ‘plus.’

Variations from ‘normal’ are noticeable phenomena. They’re noticeable by
reference to whatever it is that’s ‘normal for me.’ And it’s the fact of the
variation which is relevant to some state being noticeable, and not what the
normal state’s features are. That is to say, if you sleep four hours a night
normally, that doesn’t make how much you sleep noticeable. Two hours
might be ‘poor.’ That would make it noticeable. Six hours might be ‘poor’ for
somebody else; that would make it noticeable for them. You don’t have to
have an equivalence.

It’s also to be observed, and equally crucial to the whole business, that the
variation categories are standardized, and they’re standardized without respect
to what they contain or what the normal contains. Any Member can employ
the set of categories to formulate their current state; that is, they can say
‘normal,’ ‘poor,’ or ‘great’ without reference to what it is that that stands for,
or how what it stands for compares with what anybody else has theirs stand
for, and they can talk about it. And if the product of some monitoring comes
up with one of the variant states, that provides that that state is noticeable,
and provides, then, an occasion for an account of that variant state. That is,
it provides for an inquiry being launched as to how come it’s that. If there is
a collection of variances, then the problem that an inquiry has to solve is their
co-occurrence.

Now consider how extraordinarily elegant this is, as compared to a
situation which is imaginable if one had to have a standardized content to
these things. That is, if one had to have a standard measure such that persons
would have to decide whether ‘my normal’ was ‘normal,’ and if one had to
have measures for the variants, and it was required to know what the ‘normal’

Lecture 8

59

was, for any given person to decide that ‘my poor’ was ‘poor.’ But such
standardization is, not absolutely irrelevant, but essentially irrelevant to the
way these measurements are actually done.

There are very special cases, which are set up just by this thing being so
extraordinarily simple, where these matters do come into question. That is,
there are some physiological states which are dramatically different with
respect to what is ‘normal,’ where that might be taken as something that
ought to be found out about. But an enormous amount of diagnostic talk -
with professionals, and in conversation which is not with professionals - can
go on without much further specification than that; without having to check
out what is ‘normal,’ what is ‘poor,’ whether that’s ‘really normal,’ whether
that’s ‘really poor,’ etc.

It’s that set of facts, and the set of items for which those facts hold that
provide part of the superstructure that permits these conversations to take
place. People don’t know each other? They don’t need to know each other and
nonetheless all this can go on. With respect to these kinds of items we do not
have what we ordinarily refer to as ‘ambiguity troubles.’ Asking, for example,
“How’s your appetite?” and hearing “Not too good,” that’s sufficient. What
it is that “Not too good” is for that person doesn’t have to be known. And,
for example, that someone eats an awful lot normally and now what he eats
would still be a lot for most people, doesn’t matter.

What I’m trying to differentiate here is the following. These are clearly
measuring categories of a sort. And one might have the idea that, given the
way measuring is done, these are ‘rough versions’ of something that is
mathematicalizable; that persons are talking loosely about something which
our business would be, in studying it, to find out what the specific measures
are. What I’m saying is, there aren’t any such measures built into the use of
these categories. I’m pointing to the fact that here’s a medical device of a sort.
Its power comes from its emptiness in a way. It measures directions, and it can
work for normality apparently by reference to ‘whatever it is that normal is,’
and that’s enough. That there is a set of items like this is fantastic. Consider
what a time system would look like, or what a monetary system would look
like, built on this kind of a device.

What we want to consider carefully, much more carefully than sociologists
have ever considered the matter, is what kind of an object do you have for the
purposes of counting? Here is another way I came to think about this matter.
In about 1956, in the Journal of World Politics, Talcott Parsons wrote a review
of C. Wright Mills’ The Power Elite. I’m not a great fan of Parsons, but there
he made what I thought was a really basic kind of point. He said, “In this
book Mills seems to propose that power is a zero-sum phenomenon.” And
what that means is, if I have it, you don’t. Something adds up to a number
that can be divided among the set of persons such that if 90 percent are X,
then 10 percent are Y; that is, there can only be 100 percent.

Quite obviously there are lots of things that are not zero-sum phenomena.
For example, an economy may not be a zero-sum phenomenon; it can expand.
If you put something in you can get more out. And if everybody has a certain

60

Part I

amount, that doesn’t mean that that’s all there is. And when you’re
considering social phenomena, one thing you want to do is to try to find out
what kind of a counting system you need. Is the object zero-sum or
non-zero-sum, or something quite different? And then, how is it that persons
go about counting this or that matter?

Let me give a rather simple kind of consideration. Here’s a quote from a
group therapy session of seventeen, eighteen-year-old kids. The group is made
up of four boys and a girl. However, the girl left the group the week before.
In this exchange, one of the boys is talking to the therapist, also male.

A: Did Marian call or anything this morning?

B: Why, did you expect her to call?

A: No. I was just kind of hoping that she might be able to figure out

some way of coming to the meetings. She did seem like she wanted
to come back.

B: Do you miss her?

A: Oh, in some ways yes it was nice having the opposite sex in the

room, ya know, having a chick in the room.

The first thing to notice here is that MIR categorization device is being used
to formulate the absent person, the relevant set ‘sex,’ the category ‘female.’ In
general, it’s to be noticed that the MIR device provides one of the basic ways
that Members go about counting all sorts of things. This can be done, since
for any category-set there is a known set of categories, and what you do is
examine some population - a meeting, a party, one’s employees - deciding
whether, for this or that category-set, every category is represented. You can
have, then, a notion of completeness and a notion of absences. You can then
formulate the sort of thing you need to make this or that population
complete, where you don’t have to know the people you need; you only have
to know that they have to be ‘females,’ for example.

In this particular case, the way this fellow finds of formulating someone’s
absence is psychodynamically interesting. He is, for various reasons, much
afraid of ever showing affection for anybody. And as much as he could do was
to provide for the fact that ‘some such one’ was absent, and that was good
enough grounds for him to miss her. But he wasn’t going to say that he likes
her or anything like that.

More abstractly, such a use of the MIR device is a way one can go about
making what I think can reasonably be called ‘safe’ or ‘non-commital’
compliments. A category, in this case ‘female,’ is applied to someone, which
discriminates that one from every other person in the place, so that someone
is characterized by reference to a category that applies to nobody else. The
relevance of that for making a safe compliment may be considered if you
think of compliments like “It was nice having someone smart in the room”
where, on its use, anybody else can say “Well what about me?” That is to say,
there is a whole range of categories which also can apply to any other person
in the room, and thereby if one is singled out for notice on that item, the

Lecture 8

61

others may be implicitly derogated. Another thing is, just to mention it, that
on the category-set used to provide for the way that she’s located, it’s not
simply the case that she’s a member of a category that nobody else is a
member of, but that everybody else is a member of the same category by
reference to the one that’s been used to locate her.

Now, there are whole series of ways in which we can talk about Members’
counting, measuring, adding up, etc.; that is, we can go about mathematizing
social events. Here’s an interesting way in which this matter comes up. A
woman is talking to her husband; their son had a birthday that day.

A: I said did Grandma bring you anything for your birthday, I forgot all

about it. And Nicki said uh

B: Yeah I did too this morning.

A: Forgot all about it.

B: Oh boy. Well.

A: I said, Nicki I said Mama has so much on my mind I forgot all about

it.

What we can see here is how some set of things can exclude some other thing.
The issue wouldn’t be that one had 18 things on one’s mind and therefore by
some logical operation the 19th could not be held in mind. We need to ask
what it is that stands as excluding something, when and how any given set of
things can provide for the exclusion of another thing, or can substitute for
something - all of which are mathematical operations.

Then there is a very nice issue, that of ‘erasability. ’ Some things are in a
sense erasable; for example, one can be forgiven. But in another sense it can
be seen to remain there, on the sheet as it were, such that if that thing happens
again, someone can say “I already forgave you once.” Other things don’t
seem to have even that sense of erasability. Take, for example, suicide
attempts. For many diseases, let’s imagine that the record of your life is a
hospital log with a series of entries. Now let’s say there’s an entry: “Had
pneumonia on such-and-such a date. Cured.” That’s that. If you get
pneumonia ten years from now nobody says “Gee, how come they didn’t cure
him of pneumonia?” You can get it again and this prior one had nothing to
do with it.

So in that sense, pneumonia and many diseases are erasable. The cure
removes it from the record. And the next occurrence is not seen as explainable
by the earlier one. But suicide is not that way. If you make an attempt at some
point in your life, then the fact that, for example, you went through therapy,
or you were seven years old and now you’ve grown up, does not remove it
from the record. That thing remains alive. So that if you kill yourself 40 years
later, persons will refer back: “See? He made a series of attempts.” It’s not an
erasable matter.

One relevance of suicide not being erasable is the following. Let’s say you’re
now considering whether to go into therapy, as compared to killing yourself,
where this would be your first suicide attempt. If you just do it, and do it in

62

Part 1

a way that might not look like suicide, your family won’t have stigma, can
collect insurance, and all the rest. You’re in a situation quite different than if
you go for help. If the help fails, thereafter you can’t kill yourself in the same
way. And people know that. And that provides a real trouble about getting
people into therapy. Because if they come in under suicidalness, just by going
in they have cut off what they see as an opportunity. Because what happens
is, if you die in some way which might not look like suicide, the fact that you
had that record provides that people ought to look at this death by reference
to that record, and they may decide “Suicide.’’

If you keep suicide attempts off your record, then if you do kill yourself,
you may have something important going for you. And that is, the
procedures employed for deciding that somebody committed suicide. There
are kinds of events which are well recognized to occur. For example,
somebody can sit down and drink themselves to death. However, it’s a very
hard thing to do, and if pills are available, people, it’s said, just don’t do it
that way.

Now if you had a statistical operation for deciding did so-and-so kill
himself, you’d take 100 cases and you’d have a notion of the probability of
events occurring, so that in 5 percent of the cases where persons died of
alcohol they may have killed themselves. You could then bunch those cases
and you’d get a distribution at the end, where 95 percent of deaths by alcohol
are accidental and 5 percent are suicide. You could pick the five cases that
looked closest to being suicide and propose them as probable suicides. It
doesn’t happen that way. Cases get decided one by one. And each time a case
comes up, the fact that there’s a 95 percent chance that any death by alcohol
is not suicide, is used to decide, “No.” And you get the statistical outcome
which is that they never occur.

My own feeling about such matters is that a range of decisions are made in
terms of ‘odd events’ versus ‘normal events.’ And odd events, by and large,
are just not added together. So that if one has a notion that some X is a
normal event, then the fact that occasionally or two or three times in a row
something else happens, that doesn’t provide for a shift. One doesn’t now say
“Well, maybe X isn’t the normal event.” But, in part perhaps by way of the
fact that what is normal gets incorporated into things like proverbs and
becomes very stable, odd events are just sloughed off. They don’t get
incorporated. The fact that it’s odd is enough to mean that one doesn’t have
to consider it on this particular occasion. What you get is, “Those things
happen, sure, but ...”

One thing about odd events, then, is that they’re very hard to report. This
fact can occasion the relevance of a category that this society has, called
‘believing.’ So, for example, there are classes of events which are very closely
tied to activities that someone ought to do. If the event occurs, the activity
ought to be done. Now for some events that are so tied to an activity, those
who ought to do that activity may not be in a position to observe the event,
but they have to deal with a report. Somebody says the event happened. Fire.
Wolf. There’s a kind of gap, in the sense that one can’t wait to see whether

Lecture 8

63

the thing that stands as a condition for some action has happened, but one has
to act on the basis of a report. And of course the Cry Wolf fable is the fable
about this problem.

What’s to be done when if something happens some action has to be done,
and nonetheless the persons who have to do that action can’t see to decide?
And that’s of course where ‘believing’ fits as a category. ‘Believing’ can
operate on such a basis as the following. A notion of credit is built up about
somebody’s reporting of events. And his credit is used to decide whether the
action ought to be done. It might be that somebody has a specific task,
reporting event A, for which some action should be taken. And if they report
it wrongly a series of times, they’re no longer creditable on event A. It can also
be the case that ‘believing’ does not become event-specific, but that for some
person, whenever they say something happened for which some action should
be taken, whatever the event and its relevant action, they just don’t have any
more credit. It may be they have no more credit across whole sets of classes
of events, or it may be that while they’re not believable in general, for some
events you’d figure this is not the kind of thing anybody lies about. And if one
was interested in mathematizing social events, one would want to find out
what the classes of events are, when they overlap, etc. For example, which
kind of event is used to decide that persons are unreliable on another kind of
event, and which are not.

Now one of the things about ‘believing’ is, if you can propose that a person
has no credit, then it seems to be quite adequate to use the fact that you didn’t
believe them, even though the event they reported did occur. In my suicide
cases there are situations where a person has killed himself, investigators ask
a friend or relative or neighbor, “Did he ever threaten?’’ They say,
“Threatened all the time but never did it, so we never took him seriously.”
And that’s treated as quite an adequate account by the persons involved, and
apparently by those they tell it to.

Well, that’s fine, if what is at issue is building an adequate account. But
suppose you have the following problem - and it’s a live problem. You’re
setting up a missile receptive system, where dots appear on a radar screen.
And the people who monitor the radar screen have to tell somebody to act.
Now ‘believing’ is the resource the society has. But the one who has to act can
know that the screens can show dots when there’s nothing there, or can show
dots and there’s something there but it isn’t missiles. And if you believe and
act upon the reported dots, you’re in plenty of trouble. And if there are dots
and they are missiles, and you haven’t believed the reports, then you’re also
in plenty of trouble. But that ’s the category that people seem to be stuck with.
And of course in this case an account will obviously not be adequate in the
sense that, whether there were no missiles and you acted, or there were
missiles and you didn’t act, nobody is going to be around to say, “Well, so
you did your best.”

I’ve heard talks by persons who are in the business of, I think they call it
‘human systems’ for things like the Air Force, and as I understand it, this is
an actual problem that they don’t know quite how to deal with. It seems

64

Part 1

they’ve tried some experiments designed to select persons whose reports
would be reliable. And that would be decided by how persons behaved in the
experiment. Now that would seem to be fairly reasonable. Except that these
guys know it’s an experiment, and now they’re going about considering what
happens if they behave too conservatively, will they ever be chosen? And if
they’re not conservative enough, will they ever be chosen? And that of course
affected the outcomes of the experiments.

Now, for some sorts of organizations there seems to be a system for
deciding reliability that has to do, not with behavior but with category
membership. One day I was going through some police files on suicide and
I came across a series of notes to the police involving the discovery of a body
in a college dormitory. A fellow was found hanging in a closet. Here’s the
sequence of notes.

The first note says, “I am Mrs so-and-so. I am a housemaid in the
such-and-such dormitory. At 10:08 in the morning I was cleaning out room
472. I had swept around the room. The closet door was slightly open. I
opened it a little more, and there was somebody hanging there. I immediately
stopped. I went down and found my supervisor, Mrs so-and-so.”

Next note. “I am Mrs so-and-so, supervisor of such-and-such floor in
such-and-such dormitory. Mrs so-and-so came to me at such-and-such time
and said that there was a body in the closet of room 472. 1 went to room 472,
looked in the closet, and there was a body there. I called the head of the
dormitory.”

And then a note from the head of the dormitory. “At 10: 14 I was called
by Mrs so-and-so, the supervisor for such-and-such floor, who said that there
was a body in the closet in room 472. I took the elevator up, went into the
room, looked in the closet, there was a body there, which I then cut down. I
then called the police.”

You could figure my God if you want to talk about organizations and
hierarchies, isn’t that something? That on such a matter that’s the way it
operates. Well, but maybe it’s occasional. Maybe it just happened that time.
No so. Recurrently when a body is found in an insitution, you get that
sequence. I routinely clip the newspapers, and when you look at things like
murders, suicides, etc., you find things like this. There was a recent murder
in New York, and the report says, “The body was discovered at 5:24 a.m. by
the janitor who woke up the superintendant who then went and checked and
then called the police.” Some categories in a hierarchical organization do not
have enough rights to say that there’s a body somewhere without some other
category checking to see that there is indeed a body somewhere. Nobody
touches the body until somebody who is ‘responsible’ comes. And they’re the
ones who bring it into contact with the outside world.

Now, such a sequence is not reserved to those sorts of organizations. And
it can sometimes operate in extraordinarily tragic ways. There is, for example,
the possibility that if the fellow who hanged himself in the college dormitory
was alive when he was first found, he might have been saved. But the first
time I ever thought about this matter, the following happened. I was in a

Lecture 8

65

hospital waiting room. They came charging in with this dead eight-month-
old baby which they were trying to revive, the young mother, about 18, and
her parents. What had happened was, so the mother thought, the baby had
suffocated by a pillow over its face. Actually the baby had some disease, but
the mother thought it had suffocated. The mother and baby lived about five
blocks from the hospital. When she discovered the baby was not breathing -
she happened to walk into the bedroom and noticed this - she goes to the
phone and calls up her parents who live 45 minutes away, sits down on the
front steps, waits until her parents come, and they call the police. And the
police then call the ambulance.

These built in sequences! I’ve been interested in the possibility of
orderliness in what people do for a long time, and not known that they could
be working in that kind of order. Now, the way I work has been called
‘microscopic’ with, then, the usual sociology as ‘macroscopic.’ And it’s not a
bad distinction. But then it’s proposed that social events are not closely
enough ordered so that we can get results at the ‘microscopic’ level of
investigation. I take it that we just don’t know whether or not that is so.
Certainly there has been an argument, and certainly the statistical position has
won out. Durkheim posed the matter - which is the basis for the statistical
approach to sociology - that if you take the statistical figures on suicides, you
find quite an order. And you can study those. And construct theories. Then
he says if, however, you deal with such things as the accounts that accompany
each suicide, you don’t find order at all. They’re hastily made up, by low
grade officials, etc. But he did not in fact attempt to deal with the accounts.
And, in fact, his arguments as to why you shouldn’t do non-statistical work
were statistical arguments. And it may well be, for example, that the accounts
of suicides are closely ordered phenomena.

Another objection to the way I work is that it seems to be enormously
laborious, and sociologists are not given to doing things slowly. But one
reason I’m operating as I do is, I take it that the big problem is not that we
know that social events are not closely ordered, but that we wouldn’t know
how to describe them to see whether they’re closely ordered or not. I want to
see if such work can be done in the first place. Then we can repose the issue
of where are social events closely ordered and where are they not closely
ordered. But it’s then an empirical issue that has to be discovered. It couldn’t
be solved by an argument. And we’ll see.

Lecture 9
“I Am Nothing”

B : I’m a grown woman an attractive woman I have a

^real nice date

A: Do you have any (at same time as)

B: good looking guy for a date tonight and and I somehow I’m feeling
that I’m nothing (smiling sigh)

A: Uh huh

B: And I know nobody’s a nothing But I am. It’s like everybody

else is somebody or something and somewhere along the line I muffed
up

Suicidal persons recurrently say about themselves such things as “I’m
nothing” or “I’ve got nothing.” The question I’m going to address is, can we
describe how they’re able, properly, to so speak? Is there some way we can go
about constructing the procedure whereby ‘nothing’ is a possible product?

We might first ask what would the relevance of such a procedure be for the
persons involved? Maybe for suicidal persons it is said of them, and they know
it is said of them, that they and the project are irrational. If there were
procedures available so that they could report how they arrived at the
conclusion “I am nothing” or “I have nothing,” and the person they were
talking to could go through the procedure again and find that their conclusion
was correct, then that might provide that at least this part of their project is
logical and not irrational.

Or, for example, if to arrive at this conclusion one uses materials that are
properly to be considered in producing an activity, and handles them
properly, then another claim that is raised against persons who are suicidal can
be undercut. The claim has been put by Menninger who says about neurotic
or psychotic persons in general and about suicidal persons in particular, that
they lack a loyalty to reality. (Which suggests that ‘reality’ is, for this society,
a special category. Some scientists and philosophers might say that whatever
is, is ‘real,’ and the category ‘reality’ encompasses whatever happens.) A
person might use a procedure which is otherwise properly used to make
assessments, to arrive at the conclusion “I am nothing” as a warrant for
suicide, so as to show that they are committed to what the society holds is
important or sacred, and that it’s out of just this commitment that the project

A combination of Fall 1964, tape 7, side 2, tape 8, sides 1 and 2, and Winter
1965, lecture (6) with a brief extract from Winter 1965, lecture (4-5), pp. 1-3.
Transcriber unknown on all these.

66

Lecture 9

67

arises: “I’m not disloyal, I’m forced to conclude that I’m nothing.” And
persons recurrently do propose that the suicide possibility is becoming very
logical: “I tried to reason it out and that’s the way it looks.”

If we’re going to construct these procedures whereby “I am nothing” is
arrived at properly and reproducibly (i.e., others could go through the
procedure again and find that the conclusion was correct), a first thing we
might ask is, what are the objects by which Members in general make
assessments of their lives? We’d find that those objects are some subset of the
society’s recognized values; in particular, those which have a sanctionable way
of being counted.

Of those there is a special class of values, which are what I call
‘cumulative.’ These have two features; (1) If you can properly have them at
some stage of your life, then you ought to have them at any time thereafter.
(2) You can lose them, so that having them at one stage of your life is no
guarantee that you will have them thereafter. And the way they’re counted
provides that the fact that you had them once is quite irrelevant if you don’t
have them at the time you’re doing the count. The fact that you had them
once doesn’t count.

That differentiates between things like money and children, and things like
‘kicks.’ For example, the cumulativity of the value ‘having children’ is
recognized in the proverb ‘Children should bury their fathers.’ If you had
children at some point, you ought still to have them; if they are now dead or
if an irrevocable break has occurred, they don’t count. Money is counted in
the same way. When persons assess their circumstances they don’t feel that
they may add up how much money they earned over their lifetime, but
consider how much they have now. If they do talk about how much they’ve
earned in the past - if they don’t have any now - then they’ll talk about it
by reference to a disparity.

Aside from values like these, for which one can say “I have none,” there
are other sorts of things which are cumulative, but on which you get a yes-no
alternative; for example, being married. For such things, ‘no’ is apparently
equivalent to ‘nothing,’ and ‘yes’ at least to ‘something.’ That is, just because
a thing does not have the countable property of money does not make it
unassessable in the sense we’re examining.

Kicks are different. We can sanctionably say, “Well I had a lot of kicks.”
The fact that we’re having none now, or few, doesn’t seem relevant. There
may be other things of this sort. For example, if you do something that counts
in the society - an invention, a discovery, a contribution - then that can follow
you. At any future time, should a person who’s done such a thing try to make
the conclusion “I’m nothing,” it can be countered with “You’ve done this.”

Now, some of the assessable values are so structured that they are only
relevant at a certain stage of life. For example, a caller to this emergency
psychiatric hospital talks about himself for awhile and then says “I feel I’m
a bum. I’m just nothing.” And the fellow answers, “No, you’re not a bum,
you’re not old enough to be a bum.” That is, that’s one thing you’re not
eligible to be at this stage, so you can’t use that to count yourself as ‘nothing.’

68

Part I

So there’s a notion of a stage in life in which you’re entitled to say whether or
not you have nothing on this or that value. When persons 2 5 years old say in
assessing themselves that they’re unmarried, they’re told, “No, you can’t say
that yet.’’ That’s not anything that counts as ‘nothing’ at this point. These
things are standardized; it’s a matter of certain formal properties, that your
age has to be X before Y counts as ‘nothing.’ This is without regard to the fact
that a child, for example, could be said to be ‘unemployed’ or ‘unmarried.’
It may be true, but it’s not relevant in this society.

But to get a sense of this ‘stage of life’ feature as a formal one, we can note
that not all that long ago, a quite young child could be relevantly
characterized as ‘unemployed;’ say, before the child labor laws. And, for
example, there are old neighborhoods in New York which are partly
populated by descendants of Spanish Jews. Among them the proper age to
get married is about 15 or 16. Girls who want to graduate high school - not
to mention college - find themselves under tremendous pressure from their
families, and themselves. That is, the girls themselves talk ambivalently about
behaving quite progressively when they’re 19 years old and aren’t married —
or as being failures. And there are Jewish guys in the neighborhood, but not
Spanish Jews, who date these girls and find it tremendously troublesome
because they are too well received by the families. The fact that you’re 17 or
18 makes you perfectly eligible to get married to their daughters, and if
you’re taking them out, you must be prepared for that.

Related to the stage of life feature is that of ‘prospects.’ A person who
engages in the assessment procedure I’m describing, and arrives at “I am
nothing,” may then find somebody saying, “Well, that’s where you are now
but what about your prospects?” Or, as we find consistently for any relevant
question, this question could be expectable without ever being asked; that is,
it can be independently dealt with by the person doing the assessment. So how
do they deal with the matter of ‘prospects’? Here’s some further talk by the
woman I quoted at the start of this conversation.

B: And I do think what man wants a neurotic childless forty year old

woman No man.

A: Well I’m not

B: I know that

A: sure about that

B: I know that outside of to hit the sack with

A: Uh huh

B: That’s not what I want

A: Uh huh

There seems to be a notion of ‘entitlement,’ whereby one examines some
problematic outcome by reference to the Members’ theory of how that
outcome is arrived at. When I talked about the MIR device, I mentioned that
arrangements of its classess provide pathways which persons can use to talk
about what they’re entitled to expect; given where they are now, where they

Lecture 9

69

could come out. Such a use of that device can involve formulating one’s
present state in such a way that for the collection of categories one is presently
in, one has no legitimate expectations of ‘getting there. ’ One can then propose
that one has ‘no prospects’ on that item.

In this particular case, what is being dealt with could be called a theory of
how it is that men go about selecting a mate, and who is eligible to be
selected. And there are some very specially interesting things going on in this
statement, “. . . what man wants a neurotic, childless, 40-year-old woman?
No man.” What has to be seen about that formulation, “No man,” is that
it’s to be understood as ‘no man Member.’ That is, we’re always talking about
classes and class members. That’s important, because you can propose that
any actual man who would go about selecting you is not ‘a man.’ Or, if he
is ‘a man,’ if he claims to be selecting you, that’s not what he’s up to; that is,
he’s not up to selecting a wife, but, as she says here, he just wants someone
“to hit the sack with.” So this business of finding oneself unentitled is not
simply a way of assessing some possible future, but a way of dealing with any
currently encountered person to find either that they’re not eligible, or they’re
just using a line.

In general, the notion that you can’t get there from here has to do with the
ways in which persons arrange membership classes to do things; to find pairs,
to get appointed to jobs, to become successes, etc. Such arrangements are
taken as ways of assessing whether it’s possible that you could do that, and as
I said, it works not simply to argue that you haven’t any prospects, but also
as a way of deciding what it is that’s happening to you.

Further, if it actually happens that somebody marries you, or you get a job
or a promotion, there is a notion of ineligibility that can get used even though
it seems that you’re eligible by virtue of having been selected. And that’s the
notion of ‘phoney.’ “I’m a phoney” focuses on just that sort of thing. The
most exquisite statement I’ve ever seen of that sort was written by a woman
in the course of a psychosis which has been named ‘depersonalization.’ She
says, “. . . the feeling itself is one of unworthiness, in the way that a
counterfeit bill might feel when being examined by a banker with a good
understanding and appreciation of real currency.” 1

We might go about describing this in the following way. For any MIR
device category, there’s a set of inferences attached to it, which are ‘common
knowledge.’ Some person is a nominal member of a category, but feels that
the set of inferences that are properly made about that category are not
properly made about them. They can count down along that list of inferences
and find “I’m not this, I’m not that,” by reference of course to a rule of
relevance as to what one is or is not. That is, one is not just saying “I’m not
the sheik of Arabia” or something like that, but: “I’m not one of the things
that, given my categorial membership, I ought to be.” And that again
provides for the difficulties such persons have in getting into a relationship

1 In the lecture, Sacks does a rough paraphrase. The exact quote comes from the
research notes, and is in Psychiatry, 23 (1960).

70

Part I

because they have the notion that what will happen over the course of any
relationship they get into will involve the discovery by the other person that
each of those inferable facts do not hold with respect to them.

Let me just mention a related category, that of ‘imitation.’ I’m not using
the term with respect to actors in stage performances, but where persons
observe some activity and say, “He’s imitating.” I came across an extraordi-
narily interesting use of this category in some of the older ethnographies,
dealing with the situation of Negroes in the pre- and post-Civil War periods
in the South. Again and again I found references to the activities of Negroes
as ‘imitating whites.’ And they were characterized as being ‘marvelous
imitators.’ Such reports are very similar to the way the behavior of children
is characterized. And in that sense, then, it’s being said of Negroes that they’re
children.

Now, we want to ask what does ‘imitation’ consist of, procedurally? How
is it that some behavior is seen as ‘imitation? One of the central things that
seems to be involved is this: When one normally deals with the activities of
a Member, apparently one takes it that they have some right to do some class
of activities, and that when one engages in making out what they’re doing,
one takes it that what one sees them doing is what they are doing. ‘Imitation’
seems to involve a way of characterizing some action which somebody does
when they are unentitled to do that class of action. And if you watch the way
the Negro slaves got talked about, or the way the emerging Negro is talked
about, you can see how marvelous a category ‘imitation’ is, because it turns
out that everything whites can do Negroes can imitate, but they can’t do any
of these things that whites can do.

So ‘imitation’ becomes a category which involves the construction of a
parallel set of knowledge for those unentitled Members, where it doesn’t
happen that as they do something one finds that there is ‘the doing,’ but as
they do something one finds that they’re able to imitate. One doesn’t see that
thing which would, by reference to the category ‘knowledge and capacity,’ be
taking place; that is, one sees a child ‘behaving like an adult;’ following adult
rules of etiquette, being able to produce sentences like an adult, “talking like
a big girl,” etc.

It’s noticeable in relation to this, that if the capacities of some persons are
treated in this way, then one finds that certain sorts of accounts that can be
applied to Members in general cannot be applied to them. So one finds that
they can’t be found to be ‘responsible’ for what they do, in a non-trivial sense.
If you watch, let’s say, the way that children’s suicides are described, you’ll see
that it’s not enough to say that the child was depressed, the child got a gun
and shot itself or whatever, but there’s an added item: How it is that the child
was free at that time to do it; for example, their mother went out of the room
and then the child took the asprins. That is, part of the causal account is some
competent person’s actions which permitted this thing to take place.

I take it that ‘imitation’ is one of the basic categories one wants to focus on
when one talks about the phenomenon of ideology. Because it makes
noticeable that there are a whole range of things that these persons obviously

Lecture 9

71

can do, which are by and large not seen as things they can do in the sense of
things that a Member can do, and the addition of capacities is treated as more
things that can be imitated. And that’s an extremely interesting kind of
blindness, if you want to put it that way. It shows you the power of this
procedure because it’s apparently a perfectly consistent and reasonable way to
talk, and the materials are thus never shifted over to be seen as, “They can do
those things.” In rather more abstract terms, we can come to see a way in
which such categories as ‘imitation’ and ‘phoney’ provide us with something
very central, in that they serve as boundary categories around the term
‘Member.’

Lecture 10
Accountable Actions

I’ll start out with several quotations. 1

(1) A: Hello. This is Mr Smith.

B: Hello. I was referred to your office by Mr Jones from the

Conciliation Court, and I felt perhaps someone there would make
an appointment for somebody there to talk to me. I don’t know
what I want to say to you except I’m confused and the trouble -
Ask the questions. I can answer them . . .

(2) A: Hello Mr Brown, this is Mr Smith.

B: I was told to call down from the Conciliation Court and speak to

someone. He gave me a card introducing myself to whoever I was
supposed to speak to, and we had been recommended to Family
Service I believe it was, and they said they couldn’t do anything
until I had talked to someone at the Emergency Psychiatric Clinic
for psychiatric examination or something.

A: Why did they refer you to us?

B: Because I had tried to commit suicide.

(3) A

B

A

B

A

B

A

B

Mr Green?

Yes sir,

This is Mr Smith may I help you.

Yes sir, it was suggested by Miss Geno that I call you.
Who?

Geno. With the Epilepsy League.

Oh yes.

We have a problem with my daughter.

(4) B: Hello.

A: Mrs Gray?

B: Yes.

A: This is Mr Smith of the Emergency Psychiatric Clinic.

A combination of several lectures: Fall 1964, tape 6, side 2, pp. 5-17, tape 7, side
1, p. 1, M6, side 1, part 2, and Spring 1965, lecture 3, pp. 3-10 (transcriber
unknown).

1 Only the third fragment is actually given on the recording. The others are taken
from Sacks’ research notes.

72

Lecture 10

73

B : Yes.

A: I spoke to your daughter who was quite concerned about you and

I wanted to talk with you and see if we could help in some way.

(5) A: Hello Ronald

B: Yeah.

A: My name is Smith.

B: Uh huh-

A: I’m a psychiatric social worker with the Emergency Psychiatric

Clinic.

B: Yeah.

A: Your father called us this morning and was very concerned for you.

(6) A: This is Mr Smith and I’m one of the social workers at the

Emergency Psychiatric Clinic.

B: Oh yeah. I would like to have an appointment if I can. I’m kind

of a depressed personality, and I’ve been under psychiatric care
for- I had been at one time, and I thought possibly some
therapy might kind of help snap me out of it.

(7) A: Hello this is Mr Smith

B: Say, my husband is suicidal and, I mean, he’s attempted it about

a half a dozen times . . .

I’ll begin off by noticing that these phone calls seem to be ‘accountable
actions.’ By that I mean that by and large on the first opportunity to talk after
greetings, the person who’s called gives an account of how they happened to
make the call. We can also notice that they are ‘symmetrically accountable.’
That is, if somebody calls the agency, the somebody calling gives an account,
as in the first three quotations. If the agency calls somebody, the agency gives
an account, as in the fourth and fifth.

Now, as I mentioned once it seems to be a fact about invitations that at
some point in some relationship, invitations are the proper way that activities
get started, whereas a measure of intimacy between the persons is that they
don’t use invitations. 2 The same seems to hold for accounts of phone calls.
Persons who call up someone they are not intimate with, often construct
accounts of how that came about. In fact, the way I began to focus on these
things in the first place was coming across something I thought was
extraordinarily strange, but I guess it isn’t. It’s a phone call between a man
and woman who don’t know each other particularly well. At some point he
says:

A: Oh I was in the bio med library and I had big intentions of working

all day and they flicked the lights and kicked me out. Well that’s just-

2 See lecture 2, pp. 18-19.

74

Part 1

I was just gonna call and see if you and your husband would like to
come over.

He constructs the operation of the university as providing for how it is that
he happens to come to make this phone call. There he was, he was going to
study, then he gets kicked out, he’s left with nothing to do, and now, by
virtue of that, he comes to make the call. It seemed a little elaborate to me,
but persons do that. And at a little later point in intimacy it seems that they
don’t. They just call up and say “Hi, how are you.” And if you ask “Why
did you call?” they say “No reason, just felt like calling.”

In the calls to the emergency psychiatric clinic, some of the accounts that
persons offer involve another organization proposing that they make the call:
“I was referred to your office by Mr Jones from the Conciliation Court,” “It
was suggested by Miss Geno of the Epilepsy League that I call you,” etc., etc.
Some of the accounts have a different character, simply announcing the
trouble: “I’m kind of a depressed personality,” or “My husband is suicidal.”

Now, for the purpose of getting a hearing, and perhaps for the purpose of
getting help, there is a sense in which the two sorts of accounts are
substitutable. You can use one or the other, and either of them can work. But
I want to notice that the use of each of these has a different source; that is to
say, different search procedures would have generated the organizational
reference and the announcement of a trouble.

To call an organization and propose that the call takes place by virtue of the
operation of some other organization, involves informing those you’re now
calling that this action is not the first step in the search. That’s non-trivial
because persons tend to start a search for help with that place about which
they have the most rights to have an expectation that they’ll get help. That’s
very unfortunate in a way, because if they don’t get it there, they may apply
a formula to that fact: “If not there, where I had most rights to expect help,
then where can I expect help from?” That is, they treat the sequence of calls
as informing them about the likelihood that any next call will be a success.
And they may, then, give up. I have mentioned a few calls in which that’s
proposed. 3 In one, for example, a woman says, “I’m a Catholic and you
know, I wonder about Christianity. I just called three Catholic sources. None
of them offered anything.” In another, a fellow calls up and says, “I’m very
ashamed to be a Jew,” and goes through a report of the difficulties he had
when he called a Jewish agency. He says, “Now there’s nothing you can do.
I’m going to kill myself. This treatment I got was the last straw. I only want
it to be known why, and that’s the reason I’m calling. You should tell them
they ought to treat people better than they do.” And I talk about the way in
which persons take it that when they call an organization the treatment they
get is representative. As this Jewish fellow says, “It’s not that they did it to
me, they don’t know me, and I don’t feel personally affronted. But that they
treat people that way!” And at least some organizations have a solution to such

3 See lecture 6, pp. 43-4.

Lecture 10

75

a problem. They can come off without having refused help by proposing that
they are giving help, where the help they are giving you is telling you to call
this other agency which is specially capable of doing it.

Let’s turn now to the other sort of account; those in which persons propose
straight out that they need help. With respect to how that would have come
about, a first thing that seems to be worth considering is how asking for help
is regulated. In an earlier discussion I drew a distinction between two classes
of ‘others,’ and said something to the effect that one class were persons with
respect to whom there was such a bond of obligation that you could directly
say to them “I need help,” for example, ‘family.’' 4 We can talk of the two
classes as ‘insiders’ and ‘outsiders.’ The issues involved here are extraordi-
narily interesting, but to consider them fully I have to introduce a series of
other concepts. I will deliver a long discussion on the matter eventually; for
now, I’ll just sketch it. 5

There are strong feelings - and strong maxims which provide those feelings
- that you don’t turn to “a stranger” for help with personal troubles. And you
certainly don’t as a first act. There’s a proper sequence. And the first place you
turn is to your family, or to some insider with respect to yourself. When
somebody calls up an agency and says “I need help” - which is the proper
thing to say when you’re talking to an insider - a signal ought to go up that
something is going on here. What can be present is, they’ve turned to an
insider and gotten no help, or for some reason they find that they’re not able
to turn to an insider. And in this regard, something which might stand as just
another way that callers to this agency start off, may be seen as an alternative
to ‘‘I need help.” Things like this:

A: Hello. This is Mr Smith

B: How do you spell your name?

A: S-m-i-t-h

B: I would like to have some information about your Emergency

Psychiatric Clinic.

A: Hello. This is Mr Smith

B: How does this Emergency Psychiatric Clinic work out?

A: How does it work out?

B: I mean what do you do.

A: This is Mr Smith. May I help you?

B : Yes, I heard that you help people who are on the verge of committing

suicide or something.

4 See lecture 2, pp. 14-15.

5 The several lectures in which these matters are discussed are not included in this
volume: Fall 1964, tape 10, sides 1 and 2, tape 11, sides 1 and 2, and tape 12, side
1. For a formal version see H. Sacks, ‘The search for help: no one to turn to’ In E.
Schneidman (ed.), Essays in Self-destruction (New York: Science House, 1967),
203-23.

76

Part I

A: What was that?

B: I said that you help people who are on the verge of committing

suicide.

A: Yes, we do, certainly.

B: In what way?

That is, we get callers starting off with ‘requests for information’ which
involve, then, that over the course of the conversation help will be offered
without being ‘asked for.’ Because asking for help is something that one
doesn’t do with respect to strangers. So, “I need help” and “How does this
organization work?” seem to be alternatives in the sense that persons who use
them have perhaps come to do this call by the same paths.

Now, if a person does turn to an insider and does get help, it’s notable that
the form of help may be that the insider calls the agency for them - thus
setting up the sort of thing we see in our fourth and fifth extracts. Or the
insider may refer them to the agency. And calls made on that basis will take
a very similar format to the calls where the person says “Such-and-such
organization told me to call.” For example:

A: This is Mr Smith. May I help you?

B: Well, I don’t know. My brother suggested that I call you.

A: I see, well he must have had some reason for making the suggestion.

Has there been some personal problem or difficulty that you’re
experiencing?

B : Yes, I just lost my wife and I feel awfully depressed.

And there’s another which goes, “A friend of mine told me to call.” “Why?”
Caller says something, and then the receiver says, “She certainly did the right
thing.”

I want to stop here and show a way to analyze classes of statements for sorts
of things we can find in them. I had these exchanges, where a person calls and
says “My brother suggested that I call you” and “A friend of mine told me
to call.” I thought gee, that’s curious. And I tried to think of a paradigm line,
of which ‘X told me to call’ is an instance. It looked like the kind of line about
which a psychiatrist would say, “This is a dependent person.” And then I
tried to see what it was that made it look like a matter of ‘dependence.’ I
thought of the following. A kid comes into a grocery store and says “My
mother told me to buy a dozen eggs.” It looks exactly alike. It looks very
much alike, anyway. And that’s one of the major prototypic types of
childrens’ accounts; naming some adult who told them to do some activity.

Parenthetically, we can note that the fact that such an account by a child
is sufficient and they learn that this is sufficient very rapidly, causes rather nice
trouble. They can go from one to another among the set of adults who can
warrant some activity, to find one that will say “Okay.” And I have some
stuff in which children were watched over a period of time, and we find that
they do systematically go from one to the other without telling the one that

Lecture 10

77

they asked the other, or what response they got, so as to provide that one of
them will say “Okay.” Then they go back and say, “He said okay.” And of
course that’s known to generate fights in the family, etc.

So we get this prototype, ‘an adult said to do it,’ and something much like
it done by callers to these agencies. And what it does, of course, is to provide
that some person other than the caller - or in more general terms, the doer of
some action - is the competent individual in the case. Now, apart from the
fact that a person is proposing that they’re not a competent actor where as
adults they might be expected to be competent with respect to some item, the
importance of this is the following sort of thing. If one can formulate what a
child’s set of resources are at some point, then one may be able to examine
those behaviors of adults who are said to be ‘infantile,’ to find just what it is
about them that provides that character. By comparing a child’s and adult’s
version of similar statements, one may get an idea of what some of the tasks
of socialization are. One can then watch the development of an ‘adult’ by the
shift in the use of such blank forms as ‘I’m doing X (action) by virtue of Y
(competent agent).’ One may begin to see over the course of a child’s growth
that the child will be using himself as a competent agent for an expanding
class of activities. On the other hand, where someone is a presumptive adult,
about whom it’s said, “They’re infantile,” we might find that they are
producing a statement in which X is something that this person could do on
their own, but the Y proposed is some other agent.

Or, for example, by examining childrens’ resources we can get a sense of
what the psychotherapist is talking about when he says that neurotic adults do
not have a good sense of reality and, again, that they remain children. Let me
give a conversational excerpt.

B : How do you make people love you? How do you do it? I wish I knew

I wish I see people doing it all around me and I try to imitate
them and I don’t know how you do it.

A: Uh huh

B: I don’t know It isn’t because You don’t do it by loving them

you don’t do it by being thoughtful of them You don’t do it
by being understanding ... I I really I don’t know how you make
people love you I just don’t know

One of the things the woman is doing is asking for instructions: “How do you
make people love you?” Then she goes through this list: “You don’t do it by
loving them, you don’t do it by being thoughtful of them, you don’t do it by
being understanding.” Now where does she come by that list? They’re pretty
familiar objects; norms. They’re a set of things one is told one ought to do to
get what she proposes she didn’t get, presumably by using them.

I want to focus on things like this for awhile; they’re extremely important:
It’s rather well known that very young children have, from the perspective of
adults, a rather poor notion of causation. They don’t know how things
happen to happen. Now, among the ways that adults go about formulating

78

Part I

rules for children, are two which it’s important to distinguish. Call them Class

1 and Class 2. A prototype of Class 1 is, “Don’t stick your hand on the
stove.” Prototypic of Class 2 is “Honor thy father and mother’ - and such
things as “If you want people to love you, you should love them, be
thoughtful of them, etc.” belong in that class.

With respect to an adult’s conception of reality we would say that these
two are different, in that for Class 1 the consequences, whatever they are,
naturally flow from the act done. If you stick your hand in the fire, you get
burned. Whereas for Class 2, that’s not so. For a lot of things that you do that
are said to be wrong or harmful, somebody has to do something to you for
you to get the negative consequences. You can ‘get away with’ things of the
Class 2 sort.

Now it’s supposed, and it seems reasonable, that there’s a stage when
children don’t know the difference between those two classes. That fact is very
important for adults because they exploit it heavily. The way they exploit it
is, they’ll formulate a whole range of what, as adults, we would talk about as
Class 2-type rules, in terms of Class 1 operations. That is to say, they
formulate a whole bunch of those rules for which the consequences occur only
when somebody does something, as though the consequences occur as a
natural fact of life apart from anybody’s doing anything. They’ll do that even
though the consequences obviously involve the adult doing something. So
that parents say to children while giving them a spanking, “I don’t want to
do this, it just had to be done,” retaining thereby the relevance of Class 1.

The fact that there are these two separate classes, and that adults do not
make a large point of discriminating them, but instead assimilate the one to
the other, formulates a very serious set of problems for children. And that is
that children are repeatedly faced with the question, “What kind of a rule is
that, that I’ve been told? Class 1 or Class 2?” There’s no principled way for
them to find out. They have to proceed case by case. And proceeding case by
case, they can get into a hell of a lot of trouble. Adults know this and have
a whole class of proverbs on it, an instance of which is, “If you tell children
not to stick beans in their ears, they stick beans in their ears.” Children have
to find out what ‘sticking beans in their ears’ is a case of. Bettelheim, for
example, reports kids in Chicago who do things like get into a barrel and roll
down a hill onto a main street where there’s fantastic amounts of traffic, just
checking out whether it’s so that you can get hurt. And seriously disturbed
children are those who go about assimilating the whole range of Class 1
phenomena to Class 2. They then go about checking out the causal properties
of the world as though they were normative properties in the sense that Class

2 rules are.

When children get brought up with a certain amount of assimilation of
Class 1 and Class 2 rules, of course they cause tremendous trouble for their
parents and other adults. For one, if the child is about to do things like
sticking forks in outlets and other such things, then an adult has to be around
a lot of the time. And if an adult is around a lot of the time, and is saving the
child from these pains, then we get another sense in which the child is learning

Lecture 10

79

that the rules are indistinguishable. Whereas an adult spanking a child might
want to retain the relevance of Class 1; here the child is learning that Class 1
rules are Class 2 rules because it depends on whether an adult is around to
punish you, that you get hurt or not.

Now a major device for children, for separating these things out, are things
like lies and secrets. They begin to discover that there are some things which
they can violate, that, if the adult doesn’t know, isn’t told, doesn’t find out
about, nothing happens. And that may operate in alternative ways. One way
it may operate is that they develop an adequate notion of reality; that is, they
separate Class 1 and Class 2 rules. An alternative way that it may operate —
and this is one of the ways that a person who is said to be neurotic can be seen
to be neurotic - is that one doesn’t see rule violations which have had no
consequences as involving, therefore, Class 2 rules, but as involving Class 1
rules with no time-bounds on their operation. And one spends one’s life
awaiting the natural consequences of those actions which were violative of the
rules. Persons in such situations live out a great part of their lives under the
sense of the impending consequences of their violations of Class 2 rules seen
as Class 1 rules. That’s a considerable part of the ‘neurotic sense of guilt;’ an
ever-present sense of guilt which consists of their knowledge of the set of Class
2-type violations that they’ve done, like not loving their parents, etc., for
which they haven’t - yet - been punished.

What they tend also to do by not putting a time-bound upon the operation
of these rules; by not saying if at some time the consequences haven’t occurred
then they must be Class 2 not Class 1, is to formulate these sets of rules as
‘prophecies.’ So when their parents had told them “If you don’t change your
ways no man will ever love you,” then you’ll find forty or fifty year old people
who eventually can give as an account of the failures of their romances that
indeed they never changed their ways, and their parents were right. So they
live out their lives under these rules as prophecies of what it is that will
happen to them.

And persons can be found to hold the notion not only that the fact that it
hasn’t happened yet doesn’t mean that it’s not going to happen, but that that
only means, for example, that when it does happen it’ll be even more
ferocious. “Things look great now, but boy, wait till it happens.” And
positive things are treated, not as contradictions with what your parents told
you, but, for example, as the rise which will make the fall even more
dramatic. There are whole sets of paradigms in the history of the culture
which stand as examples: Sodom, the decline and fall of Rome, etc., where
the enormous rise can be treated as an example, not of a rise which inevitably
happened to fall, but the rise that was so great to make the fall worse. There’s
a thing that the therapists who are lucky enough to have such patients - lucky
enough because the patients are all rich - call a ‘success syndrome.’ And that
involves, let’s say, a man who comes from ‘nowhere,’ ‘the wrong side of the
tracks,’ and at a fairly early age amasses enormous success - in business most
particularly. And at that point when he’s been elected executive vice-president
or whatever, he goes into a depression and becomes suicidal. And what it’s

80

Part I

made out to be is the fear that he’s finally arrived at that point where the drop
will be appropriate to the kinds of sins he’s committed. The sins were such
that it wouldn’t be sufficient for him to lose some trivial $ 50 -a-week job, but
now he’s being primed; fattened as it were, for a more appropriate slaughter.

So that just gives, I think, a sense of what it is that the psychotherapist is
talking about when he says that neurotic adults do not have a good sense of
reality. And also when they say that they remain children, and that the
projected operators in their presumptively adult lives are always parents, you
can see what it is that the psychotherapist is seeing.

Lecture 11

On Exchanging Glances

An assignment has been given, having to do with observing people exchanging
glances. The recording starts after whatever description of the assignment has
been made.

Q : Can I ask a question about this assignment. 1

HS: Yes.

Q : For each person that we notice looking over at somebody are we

supposed to ask them if they know the person?

HS: No! Don’t ask!

Q : But we have to write down the class and everything, how can we

possibly know if we don’t ask.

HS: Yeah, class membership doesn’t mean Junior, Senior.

Q: Oh I realize that.

HS: I don’t mean social class, either. I mean class in the sense that I’ve
been talking about class. Any class. Whatever it might be. You
figure it out. When two persons exchange glances, see if you see
anything similar between them, and see if you see what it is that
might be what they’re noticing. If you find that you don’t know at
all, you can say that. I don’t think you’re that naive. You walk
through the streets and you’re constantly classifying the persons you
see.

Q: ((re would it matter if the glancers knew each other; e.g., were sitting

at the same table and someone walks by and the two at the table
exchanged the glance.))

HS : No, I said I wanted persons who are not interacting with each other.
Though you can add variations if you please. But the specific
assignment is for persons who are not otherwise interacting with each
other.

A combination of three sessions. On tape 6, side 2, prior to the lecture proper, an
assignment is given and discussed. Then on M4, side 1, part 2 an entire session is
given to remarks based on the papers handed in by the students. That is continued
on tape 14, side 1.

1 This lecture differs from the others in that, among other things, it includes many
student comments, contributions, and questions. These have been omitted from the
other lectures as much as possible, materials generated by them having been
assimilated into the text.

81

82

Part I

I’ll give you an example of what I mean, just from a lay approach to the
matter. I was walking down the hall the other day, to give an exam to one
girl. She was standing, leaning up against the wall. In between us walked
another girl. She passed this girl first, and then me. And the girl who was
standing leaning against the wall looked at me and gave a shrug of her
shoulders with a big smile, which I returned. And I don’t think it was a big
puzzle over what was going on. The girl who walked by was smoking a pipe.

Now, the two of us knew what we were noticing. But that can be
problematic. For example, on the Berkeley campus or in places in Berkeley,
you often find interracial couples wandering around, one of whom is Negro,
one of whom is white. And people who look like tourists, visitors to the
campus, etc. - that is, strangers - will stop to look at these couples, and then
check out with others around. The question-form might be seen as something
like “Am I in Rome, or am I here?” That is, Rome, generically. “Do I know
where I am so that I know that that’s something okay or something odd.”
And when they do that, people will not infrequently just look back at them
and give them a negative stare. As if to say, “Who the hell are you.” And
that’s treated as very disturbing.

So sometimes when you search out somebody to exchange glances with, it
may be about something you took to be deviant. And for something deviant
it may be that you look at anybody. Now, for other noticeables, persons may
pick out special classes. I want you to try to see what the classes are. I have
by no means collected these things systematically, and it’s done all the time.

Q : Something occurs to me. I can usually tell when some sort of

interaction is occurring, let’s say, and someone is looking at me. But
if someone is looking at someone else, it’s going to be kind of hard
to differentiate between people that just happen to be looking at each
other, as opposed to ones that there’s some reason for an interaction.

HS: Try it and see.

Q: and if you make a mistake, you’re liable to come to a lot of false

conclusions.

HS: Try it and see. If you feel doubtful, put down that you’re doubtful.
I spend a lot of time watching people watch people and exchange
glances. And it’s often no big deal to see that people are exchanging
glances. They turn at each other, across a room or from a distance,
and exchange big broad smiles sometimes. In some places there’s
enough turnover so that these exchanges will happen pretty fre-
quently. And it seems to me that people are noticing all the time. I
may be all wrong. Maybe you can’t tell when people are looking at
each other, but you can only tell when people look at you. Maybe
you couldn’t even tell that. But I would find that odd.

Q : Do we have to watch others watching each other? or can we watch.

HS: No, I want you to be watching others watching each other. You can
add personal remembrances if you want to, or you can begin to
record the encounters happening to yourself. But as the assignment,

Lecture 11

83

you’re to be watching others looking at each other. And I’m willing
to make a fair bet that you can guess, seeing somebody get up and
start to move out of a place, whether they will be somebody that
others will notice. And you can probably say who will notice them,
knowing nothing about the persons except what they look like; both
the person who gets up and the other persons around. Because after
all, when persons look at somebody passing, they know from having
scanned the room in the first place, who to turn to to get an exchange
of glances.

Q : Is it okay if we write down, if it happened in different places-

HS: Yes. Put down the place and the time. The more detail the better.
And if you can type these it would be a help, since I’m perfectly
willing that anybody in the class takes the collection of them and sees
what they look like, writes a paper on it, whatever else. We’ll have
the product of a bunch of persons going out independently. Now
we’ll see if they look alike.

I’m going to start to talk about your first assignment. I don’t have all that
much to say yet. I find them something to think about, and I want to begin
with something that was suggested by some of the papers. Let me make a
couple of remarks about the problem of ‘feigning ignorance.’ I found in these
papers that people will occasionally say things like, “I didn’t really know what
was going on, but I made the inference that he was looking at her because
she’s an attractive girl.” So one claims to not really know. And here’s a first
thought I have. I can fully well understand how you come to say that. It’s part
of the way in which what’s called your education here gets in the way of your
doing what you in fact know how to do. And you begin to call things
‘concepts’ and acts ‘inferences,’ when nothing of the sort is involved. And that
nothing of the sort is involved, is perfectly clear in that if it were the case that
you didn’t know what was going on - if you were the usual made up
observer, the man from Mars - then the question of what you would see
would be a far more obscure matter than that she was an attractive girl,
perhaps. How would you go about seeing in the first place that one was
looking at the other, seeing what they were looking at, and locating those
features which are perhaps relevant?

Now the matter is important because there are many occasions on which
you might want to feign ignorance. I want you to come to see that that has
to be learned as a special task. Here’s the sort of thing I mean. In a short story
by Conrad Aiken which I forget the title of, some fellow is engaged to some
girl who claims that she’s exceedingly innocent. And what bothers him is that
her body doesn’t seem innocent; that she knows what to do. She knows too
well how to behave, it seems to him, to have not done elsewhere what she’s
doing with him. And the task can be quite delicate. For example, sometimes
women can be in a group of persons one of whom is their husband, and their

84

Part I

husband takes it that they’re not very knowing. And they play that with him,
but can perform in such a way that others will know that that’s not so.

Here’s a related sort of issue. I once glanced at a paper concerned with
teaching actors how to learn to make an error in doing something. For
example, if your job is to be some professional, and in the play that
professional makes an error, you have to learn how such a one, who knows
how to do something, could do that erroneously. And the various set of things
that you go about doing erroneously may give away the fact that you know
just what you’re doing. There could be erroneous ways of doing integration
in calculus, one of which would indicate that you have no idea what those
symbols are, others of which might indicate that you know how to do
integration.

I think of a classical story, a very fantastic event that happened in the last
couple of years. There was an automobile accident outside of Moscow. One
of the victims was the leading theoretical physicist in the Soviet Union, and
one of the very few leading theoretical physicists in the world. He was killed.
But they didn’t want him to be killed, and so, through some unbelievable
exercises they organized a staff from around the whole world to bring him
back. And they reassembled him, essentially. He was absolutely destroyed;
clinically died five times in the first ten days, and they had a fantastic team of
persons coming in to put him together again. Now the question was, was he
going to be the same guy? So as he began to recover - and that meant very
dramatic kinds of developments from essentially infantilism - his colleagues
would go in and talk to him.

One crucial day, one of the colleagues comes out very depressed and says,
“He’s not himself.” And they said “Why?” He says “Well, we were having
this discussion about some problems and I asked him a question, and he came
back with this answer, and that’s just not so, and anybody knows it’s not so.”
And they all sat there bemoaning the fact that it looks like he’s finished. Then
it suddenly occurred to somebody, “My God, look at that answer!” It was a
far more elegant solution to the problem than anybody had thought of. And
it was just not available to any of them, but they knew that that’s the way he
would work. So that situation was transformed very dramatically. And in that
regard, let me just note that, for example, things that children say, which they
have no right to say, can be treated as ‘errors.’ That is, errors are socially
stratified in many ways.

Let me go at this business of 'inferences’ in another way. The problem that
I am stuck on at this point, and I don’t have anything like a solution to it, is
how does a glance become an action? What kind of a world do you have to
build to make a glance an action? Let me start off by reading you a quote from
an extremely important book, the title of which I also forget. I think it’s called
Sight and Sense , but you can easily find it given that the author’s name, Von
Senden, is not that common, and he only has one book translated into
English. It was written in, I guess, nineteen twenty something. It by and large
reports on others’ investigations which had never been brought together. The
book is exceedingly important in a lot of ways. The very eminent psycholo-

Lecture 11

85

gist, Head, made a great deal of his reputation on his analysis of the materials
Von Senden presents, which Head was able to do because he knew German
and the book hadn’t yet been translated into English. And the theory that he
developed is built largely on Von Senden’s work. So far as I know, sociologists
have not used the book; the place where you would expect to find it among
the classical sociologists is Sandeke, but I don’t see that he even had any
reference to that kind of material.

The book is about persons who, typically, were blind from birth and who
got sight, and what kinds of things they could learn to observe afterwards,
and what kinds of things they never learned to observe if they got sight too
late. On pages 61-62 we get a report by Dufau of “a girl who only
discovered at the age of 12 that she differs from other people in lacking a
sense, and who now seeks to discover the nature of this unknown sense.’’ And
she says:

I posed myself a host of questions about this new and unknown state
which had been described to me, and did my best to come to terms with
them. In order to satisfy my doubts I had the idea of trying a strange
experiment.

One morning I again put on the dress which I had not worn for
some time because I had been growing so rapidly then from month to
month, and thus attired I suddenly showed myself at the door of the
entry room in which my governess was already working at the window.

I stood listening. “Good heavens Lucy,” she said, “why have you put
on that old dress that only reaches to your knees?” I merely uttered a
few idle words and withdrew. This was enough to convince me that,
without laying a hand on me, Martha had immediately been able to
recognize that I had again put on the dress that was too short. So this
was seeing.

I gradually recounted in my memory a multitude of things which
must have been daily seen in the same fashion by the people about me,
and which could not have been known to them in any other way. I do
not in the least understand how this happened, but I was at last
persuaded. And this led gradually to a complete transformation of my
ideas. I admitted to myself that there was in fact a highly important
difference of organization between myself and other people.

Whereas I could make contact with them by touching and hearing,
they were bound to me through an unknown sense, which entirely
surrounded me, even from a distance, followed me about, penetrated
through me, somehow had me in its power from morning to night.
What a strange power this was, to which I was subjected against my
will, without for my part being able to exercise it over anyone at all. It
made me shy and uneasy, to begin with. I felt envious about it. It
seemed to raise an impenetrable screen between society and myself. I felt
unwillingly compelled to regard myself as an exceptional being that
had, as it were, to hide itself in order to live.

86

Part I

Most of Von Senden’s work concerns the perception of patterns. If persons
are blind from birth and they get their sight as adults, it turns out that certain
patterns just can’t ever be learned. They may not ever learn how to see a
rectangle. They know what a rectangle is, and they can find it by counting the
points. But where one sees, quote, at a glance, a rectangle, a triangle, and
other such things, they never learned how to do it. Now there are lots of other
things that people see, quote, at a glance - and perhaps that blind persons
know the categories of. And how it is that persons come to see these things
that they do see in this fashion we don’t know very much on. But what’s
important in the first place is to try to determine, in part, what it is that is seen
in this fashion.

And that’s why it’s sort of a pain to intellectualize this stuff such that you
already talk about it as though, “I see a blob, and then I infer that it’s my
mother because she’s a blob like that,” when what you see is not that. And
it’s an extraordinary experience when it turns out that you do see somebody
in something like that fashion. I just recalled, in Fitzgerald’s novel The hast
Tycoon, he reports a scene where this girl remarks that walking in New York,
seeing a man approaching, finding a whole series of properties of him which
she doesn’t especially like, it turned out it’s her father. That is a special sort
of experience, but I can understand why you talk that way. You may find that
some philosopher has convinced you of this. But ‘inference’ has another kind
of use.

So where are you going to start to try to build a way of dealing with this
stuff? We start out with the fact that glances are actions. That’s the first fact.
There’s a beautiful report to this effect. A guy is looking at a girl, looks
around to find somebody to exchange glances with, catches the eyes of another
girl who looks like this one in some way, and turns away quickly. That is, he
sees her sanctioning his looking. And how do we start to provide for glances
as actions? I take it we have to start building classes. Earlier I was talking
about persons being representatives of classes of which they were members.
And the sorts of things you’ve reported suggest that what persons see is a class
member, for whole collections of classes. You see ‘a girl,’ ‘a Negro,’ ‘a
such-and-such.’ That is, the class permits you to see what it is that’s there. It
permits you to see.

To liven this matter up, I’ll read you something from The City of Plains
by Proust. And Proust is an incredible sociologist, as you may know if
you’ve read it - and if not you certainly ought to, even if you’re not
interested in literature. There is a scene where Proust is watching events
take place in a courtyard below. He sees a whole sexual confrontation
between two guys, which he describes absolutely fabulously. Then he
writes:

From the beginning of this scene an evolution in my unsealed eyes had
occurred in M. Charleaux. As complete, as immediate as if he had been
touched by a magician’s wand. Until then, because I had not under-
stood, I had not seen.

Lecture 1 1

87

The vice - we use the word for convenience only — the vice of each of
us accompanies him through life after the manner of the familiar genius
who was invisible to men as long as they were unaware of its presence.
Our goodness, our meanness, our name, our social relations do not
disclose themselves to the eye. We carry them hidden within us. Even
Ulysses did not at once recognize Athena. But the Gods are immediately
perceptible to one another, as quickly as like to like. And so, too, had
M. de Charlieux been to Chiupien.

Until that moment I had been, in the presence of M. de Charlieux,
in the position of an absent-minded man who, standing before a
pregnant woman whose distended outline he has failed to remark,
persists, while she smilingly reiterates “Yes, I’m just a little tired now,”
in asking her indiscreetly, “Why? What is the matter with you?” But let
someone say to him “She is expecting a child,” suddenly he catches
sight of her abdomen and ceases to see anything else. It is the
explanation that opens our eyes.

So, the classes and their categories permit you to see. That’s a start. It’s not
enough to make a glance an action. As some of you noticed and reported, it’s
not merely that some observer is seeing by reference to some category, but that
the one being observed sees what the observer is, and is seeing. And we get
into that whole jumble there: A, seeing that B is looking at A, sees what B is,
and what B sees, etc., etc.

That A can see what B is, and what B is seeing, may seem the most trivially
obvious fact. But it was the economist Bagnanin’s seeing that fact that
provided for the modern revolution in economics. Until the book A Theory of
Games in Economic Behavior, the major theory of economics used the Robinson
Crusoe model: A man alone in an environment, and now he has to go about
deciding what to do, what things will work, etc. And any other person is to
be conceived as a part of the environment. That meant that you could give
statistical treatment to the various parts of the environment. Now Bagnanin
was, among other things, a great poker player. And he saw that economics
could not be constructed along the Robinson Crusoe model, and took poker
as a model. And in poker you can’t treat the other person as a statistical object,
but as somebody who, whatever strategies you might employ to deal with
‘that piece of the environment,’ does the same about you. And then,
furthermore, knows you do use strategy, etc., etc. If you just read the first
chapter of the book, you see that laid out. As I say, it provides for a complete
reconstruction of the way of doing economic theorizing. It’s a very curious bit
of history that Parsons, in his Structure of Social Action, posed essentially the
same facts - and he left economics because of a similar complaint against
economics. But the sociologists apparently didn’t appreciate what Parsons was
posing, and have come to this kind of position through Bagnanin and
Morgenstem, and not through sociology itself.

So, A can see what B is, and what B is seeing. How is it that, given there’s
a whole bunch of classes available, A can see that? I have to introduce a notion

88

Part /

that will help this problem along. Let’s talk about there being, for some
collection of classes, an ‘order of relevancy’ with respect to categories. It seems
that a set of circumstances can provide that order of relevancy for some
membership class. If the circumstance is that A is being looked at by B, that
in itself might inform the consideration of what order of relevancy is operating
here. And if there is an order of relevancies, we can begin to locate certain facts
that obviously do occur. For example, ‘failures’ can be located that way. And
‘absences.’ People talk about “missing something,’’ “not seeing something.”

Now a third thing may be very extendable, but for now I’ll only deal
with classes that have two subclasses. For some classes. A, being observed,
sees that B, doing the observation, sees that A is a member of some contrast
class such that A is one part of a paired class and B is the other. Where for
each, there is an order of relevance which provides for the other as being an
observable, and - and this is absolutely important and the thing would not
work unless it was present - the order of relevance of each was available to
the other.

With that, we can begin to handle some rather nice kinds of events; you
can begin to see what the following kind of trouble is. A girl looks at a guy.
And what he does is, he takes her glance and then looks over his shoulder.
And what’s going on is that he knows enough to know the sort of thing she’s
looking at, by reference to the class - she’s looking at ‘an attractive
male’ - and he has that sort of insecurity which provides that he’s not eligible,
so it must be somebody else. This complementarity of orders of relevance
permits us to see how persons are operating when they talk of themselves as
‘worthless’ or as ‘nobody.’ And it’s this complimentarity - that one knows
what the other classes do - that begins to tell us what may be going on, in
part, where some guy is looking at a girl, now looks around and catches the
eye of another girl who is somewhat similar to the first, and turns away. That
is, he does not seem to just continue looking around as though, “Well, she’s
not a male so how would she know what I was doing?” Or, for example, some
of you report a person will be walking, sitting studying, just doing nothing,
and then see somebody looking somewhere, and, quote, follow the glance,
knowing that there would have to be something at the end of the glance
which is worth looking at. And if you can see what it is that is doing that
looking, you could have a pretty good idea of what it is that would be at the
end of it. So this complementarity is equally as crucial as the fact that one is
able to see what somebody with whom you are a member of a class in
common is seeing when they look at you, or another. The sense of there being
‘a society’ is that there are many whomsoevers, who are not members of this
or that class, who are able to see what it is that one is looking at.

Now there are other things that you talked about, which are worth
thinking about. For example, people talked about the fact that persons didn’t
exchange glances over an average person. Now what you have to attend is the
notion of ‘average’ that you employ. Because I don’t think it’s the case that
you use it in the following way: You’re sitting someplace, a set of persons pass
you by, and you construct a distribution over that set of persons and provide

Lecture 1 1

89

that those that stand out from that set, after let’s say, watching awhile, are the
unaverage ones. Rather, somebody’s being outstanding may be quite irrele-
vant to the collection of persons surrounding them in the scene; that is to say,
those who have passed before or who come after.

A while ago I gave some examples of odd suicides, and I said then that
given the procedure that’s used - that is, case by case - what you have is that
an odd suicide gets separated off and is not used to consider the, quote,
normal ones. 2 So a stream of odd ones is only something to be remarked
upon, like “Isn’t it amazing today, there’s such a stream of absolutely
beautiful women who walked through such-and-such a place.” It isn’t that
one then finds that one is going to modify the notion of ‘average’ because of
that. And that then means that those categories have to be given special
attention.

In many of your reports it’s proposed that what makes something
noticeable is that it’s an ‘incongruity.’ I’ve given a glancing attention to the
phenomenon of incongruity for a long time, but it’s only now that I feel that
I’m beginning to get a handle on what I might be able to say about it.

A first thing we need is a notion that what Members see is decomposable
by them. That Members can decompose some event, situation, complex,
whatever you want to call it, is no surprise at all, given the sorts of things
we’ve been considering. That is to say, we’ve been talking about activities as
being ‘assembled.’ And if that’s so, and if Members can see that that’s so,
then, that they can take them apart would not be especially surprizing. The
question then is, having taken something apart, how do they put it back
together again so as to find what it is that’s strange.

The way they seem to do this involves treating something that they see as
a combination of parts, some of which have names. And to those nameable
parts are affiliated standardized procedures for producing those objects in
some combination. I’ll give you an example. You see a man in a car. Now a
car is a ‘possession,’ and there are ways that one properly, normally,
legitimately, expectably comes to have it. And apparently what goes on is that
when one sees ‘an incongruity,’ one of the things one sees is that it does not
seem that the proper procedure for this possession having come to be
possessed, has produced the combination of this person and that car.

A guy I know told me the following story. He looks like a bum, is usually
unshaven, wears very tattered clothes, and has a big flashy car. He’s driving
down the Massachusetts Turnpike and there’s a police car sitting in the grass
that divides the highway. He passes the police car, drives along below the
speed limit, the police car starts to follow him, follows him along for a while,
finally pulls him over. The policeman asks him for his license, the guy shows
him his license. And now we get the following interchange. “What do you

2 See lecture 8, p. 62.

90

Part I

do?” “I go to school.” “Where do you go to school?” “Harvard.” “That’s
very nice. Where are you going?” “I’m going home.” “Whose car is it?” “My
mother’s.” Then the policeman says, “Look, would you do us a favor?”
“Sure. What do you want.” “Would you get a shave? if you don’t, every
policeman on the Turnpike is going to have to stop you.” That is to say, for
such a person, how it is that they come to have gotten that car is a problem.
And the incongruity is seen in that way; by virtue of the car being a
possession, how would it come to have been possessed by such a one?

Or, for example, I talked about a woman who proposed that she was
‘nothing,’ and said “What man would want me?,” given her set of features.
And in analyzing that, I said that what she was proposing was that given her
features, the correct procedures for somebody selecting a woman would not
arrive at her being in some pair. 3 And we can see a sense in which, whether
in any given circumstance the person who says that is correct or not, they are
in fact employing a procedure which might be employed if they happened to
be together with someone; that is, people would say “How did she come to
get him?” That is to say, one thing which seems to be observable as an
‘incongruity’ involves two persons being present, a man and a woman, where
there’s some special difference between them: A worn old man and a very
young pretty girl, or the reverse. And now they get noticed, an exchange of
glances takes place, and the question that seems to be asked is “How did he
get her?” or the reverse. What one is doing is employing the procedure by
which persons properly come together and finding that that does not produce
these two persons as a pair. And one can then produce an explanation; for
example, “He bought it” or “she bought it.”

So that the procedures whereby persons come to be in some combination,
or come to have some object, seem to describe in part what it is that you’ve
observed, and what I had observed earlier but had never really been able to
figure out - how it is one sees an incongruity, and also sees the possible
illegitimacy of some combination.

Now, in some of the reports what we seem further to find is that where
persons are concerned, units of larger-than-a-member are observable. People
can see ‘a family,’ for example. That may be trivially obvious, but it’s very
important, and it has to be achieved in some way. One of the reports has a
powerful instance of the relevance of persons seeing such a thing as a ‘family.’
One of you reports that you were driving along and a car pulls up and stops.
It doesn’t seem to stop anywhere special, just pulls up and stops on a street.
The door opens and a girl of about 18 charges out, runs across a lawn and
stops, and starts shrieking. In the front seat are an older man and woman. The
guy jumps out of the car, charges across the lawn, comes up to the girl and
gives her a smack right in the face. At which point some of the passing cars
slam on their brakes, and some people start getting out of their cars.

The man and the girl stand there, face to face, screaming at each other, and
then he just grabs her and drags her back to the car. And people look at each

3 See lecture 9, pp. 68-9.

Lecture 11

91

other, shrug, and say “Well,” get back in their cars and go on their way,
taking it that it’s not after all a kidnapping scene or an attempted murder, but
it’s ‘a man and his daughter’ and he’s punishing her for something she did,
or something to that effect. No attempt at intervention is made. Whereas I
take it that were the unit ‘family’ not available to them, and relevant features
of it not observable in the scene, then it could be expected that people would
intervene. But given what they saw, intervention becomes ‘meddling.’ And
one wonders, again in part, how much of the failure to give help in dramatic
scenes is a product of some order of relationships where the combination of
persons and what they’re doing is something seeable as ‘their business.’

Ql: I read in the paper where a girl was in a car and she had parked a

couple of blocks from her home. And another car pulled up and the
guy driving asked her for information, “How do you get to the
freeway?” She started to tell him, still being seated in the car. The
guy said “I can’t see you, come around and show me on the map I
have with me.” And she immediately got frightened, rolled up the
windows and stayed in there. And he said “That won’t help you,”
and he tried to get in. And she started honking her horn, hoping that
someone would come out, hearing the horn. And so what he did was
honk his born back. And therefore it sounded more like a game, and
no one came to her aid.

Q2 : The was another situation that was in the paper last week where a
woman was sitting in the car waiting for her daughter and her date
to come out from a show. And this man got into the car and started
beating her up. And she yelled for help, and there was a couple
watching, and another man. And no one would help. And after-
wards, when the guy left, she said “Why didn’t you help me?” And
the woman in the couple said “Well we thought you were married
and it wasn’t any of our business.”

Q3 : But then on the other hand there was a story in the paper about this

young motorcyclist who stopped on the freeway, who looked down
below, evidently at another freeway portion where a car had stalled.
And he leaned over the railing to ask if this man needed any help,
and the answer was a shot from a gun, and he got killed. So these
are the kinds of things that may be part of people’s fear.

Q4 : Which all points to a kind of a pattern of three things: A normal way

of going about things, a subversive way of going about things with
a normal appearance, and an uncertainty feature operating with
regard to all events, since it’s known that persons can be acting in a
subversive fashion and appear to be acting normally, at the same
time it’s not known whether in fact they happen to be operating in
a subversive fashion. And as a matter of fact it would be more of a
problem to account for why that uncertainty factor is not operating,
than to account for why it is operating when it is.

HS : Well, that may be. The problem about what you say is, in part, that

92

Part I

in this society there is a considerable shift of the responsibility for
locating subversives, from whomsoever to a special group, the
police. And persons otherwise feel that they’re entitled to suppose
that people are what they look like. And if they’re not, then that’s
not their problem very much. It’s fairly special that people use a
supposition that somebody is not what they seem to be.

Q4 : It’s Goffman’s argument, the notion that persons feel that if some-
body is acting in a particular fashion which they may doubt as being
an example of such-and-such, then they’re entitled to inspect him for
other features that they assign to a such-and-such. I think that’s much
more common than the other. Because they’re engaged in the prob-
lem of attempting to make out persons and what they’re like, and if
they now do something that you didn’t figure on, it’s obvious that
they did something. I mean they did it. So it’s obvious that

HS : Yeah but the fact that they did it is not sufficient. The first thing that
we find is that events are decomposable. And the question is, “Could
somebody have done that?” and it could be decided that they
couldn’t have done it properly, and now we get all these incongruity
observations. Now the fact is, Goffman talks about incongruity but
he does not tell us what incongruity is. That’s what I think I’m
beginning to see here in this stuff. How it is that one sees it. He has
not analyzed how it is that you do ‘an incongruity,’ what makes it
an incongruity. And l think I have the beginning of how it is that
you do it.

Now such issues as its import for integration in the social structure, which
is what you’re talking about when you raise ‘uncertainty,’ is another matter
altogether. Not to say that it’s not worth looking at. And the question of
uncertainty is in part handleable in the observations you’ve made in these
reports; that is, that people check out the things they’ve noticed with
somebody else. What we have to consider is, what is it that they get out of
checking out their noticings with somebody else? Some of you talk about
‘reinforcement;’ one reinforces their determination that someone was doing
something wrong, or that a combination was wrong, by making a check-out
with somebody else. And thereby the norms are reinforced. Which is a nice,
Durkheimian kind of argument, and it may be true. But there’s something
that has to be seen about that claim, and that is the mechanism of the
procedure in the first place, which is what I’m interested in.

In the first place, if one were engaging in some device for getting
reinforcement, it might be supposed that one would want to know the status
of the person one uses to get reinforcement from. That is, if you have a
question, “Is it the case that this thing I’ve seen is a violation of the norm?”
you might write a letter to Dear Abbey, or ask a priest, or ask your parents,
or ask somebody who has special rights. In any case, it might be presumed
that you’d be concerned to check with somebody who you knew to have some
information about it. But in your reports, apparently what we find is that it’s

Lecture 11

93

pretty much an ‘anybody’ who can be turned to for that check-out; ‘anybody’
as long as they’re a member of an appropriate class, so let’s say, a woman
turns to a woman, or a man turns to a man. You don’t ask for their credentials
in the first place, and if they return your glance, give you the smile back that
you give to them, or the disapproving glance, that seems to be okay. And
that’s a fantastic kind of simplicity.

And lurthermore, it might be supposed if one was getting a reinforcement,
that if the person did not return the same glance, give the same thing back to
you, then you would say “Maybe I’m wrong,” figure that you’re having a
wrong response. But no, that doesn’t seem to be the case. Given what you’ve
reported in your papers, what seems to be the case are several things. First, if
they don’t return the same glance, a smile for a smile, etc., you look for
somebody else, whereas if they do return the glance, you feel satisfied and look
no further. And second, if they don’t return the glance, you may make an
assessment about them: “Who do they think they are?” or “What’s up with
him?”

So the norms don’t seem to be in doubt; that is, the norm that provided
for the incongruity that provided for the noticing is not something that seems
to be held in doubt. And the notion that it’s ‘reinforcement’ that you get is
somewhat obscure. Though the question certainly has to be asked, what is it
that people are doing when they exchange these glances? We’ve already
observed that they get exchanged with whomsoever; there’s no special
credentials involved, except class membership. Another, equally extraordinary
part of this - and why, again, ‘reinforcement’ is a tremendous gloss or
oversimplification - is that one takes it that they know what it is that you saw,
that they saw the same thing, and they know what you’re smiling about, and
they make the same assessment. Their smiles tell you that. Now, that that
gets done would seem to involve a fantastic kind of social integration. And
it can give us a way to talk about ‘alienation.’ Because one of the senses of
alienation would presumably be that you would feel tremendous doubts
about doing that. And one would have tremendous doubts about what
those smiles and disapproving glances are about; what they’re looking at
you for. Schizophrenics are always reporting that people are looking into
them, talking about them, know what they’re thinking. And you can see
that they haven’t constructed a machine which produces events going on in
the world, wholely out of their heads. For that thing which they propose is
going on by reference to them, while it may not be going on by reference to
them, is indeed going on, enormously, and how it’s going on is an absolute
mystery.

The sense of alienation with respect to a schizophrenic would be, in part,
that they can’t tell when it’s being used with respect to them. Or, that they
think it works better than it works. Because it’s perfectly routinely the case
that persons know what others are thinking - without knowing those
others - in just the sense that ‘thinking’ is used as a non-technical term.
People seem to know what others are thinking without having any idea who
they are, apart from their class membership. Now schizophrenics claim that

94

Part I

others, who don’t know them, know what they’re thinking. Routinely. Well,
there’s a sense in which others do know what others are thinking, routinely.
Again, then, a schizophrenic may be wrong as to when it occurred, but the
notion they have is obviously correct; is obviously so.

Q: It seems like that was the basic assumption of the assignment; that

we could notice them doing it.

HS: Right. It was the basic assumption, which could have been wrong.
But of course it’s not like I came into this world the day before I gave
that assignment. I know that people can do this, I’ve watched it
many times, and I take it that you’ve seen it also. However, people
who get this far in their education are very prone to intellectualize
the whole operation and not see it. And so you don’t see that people
know what other people are thinking; you figure it’s a philosophical
impossibility, or you’d have to go through four years of analysis, etc.
It was an assumption. And it could have been the case that
everybody came back and said ‘‘No, I never saw that happen.” And
that’s possible. It might be something that’s dying out. A thing that
our forefathers had. Like God.

Lecture 12

Sequencing: Utterances, Jokes,
and Questions

For the linguists, almost exclusively the largest unit of investigation, the
largest unit they seek to describe, is a sentence. So grammar is directed to
providing rules for generating sentences, and every time you have a different
sentence the grammar is to be reapplied. If we want to study natural activities
in their natural sequences, we have to deal with, for example, the obvious fact
that a sentence is not necessarily a ‘complete utterance.’ Thus, linguistics is not
sufficient, at least so far as it’s by and large done. There is one major exception
and it’s extremely close to what I’m trying to do. That is Fries’ book The
Structure of English (1952).

We want to construct some unit which will permit us to study actual
activities. Can we construct ‘the conversation’ as such a unit? Can we in the
first place make of it ‘a unit’ - a natural unit and an analytic unit at the same
time? The question then becomes, what do we need, to do that?

First we need some rules of sequencing, and then some objects that will be
handled by the rules of sequencing. Now, if we restrict our attention at the
beginning to two-party conversations, then we can get something extremely
simple - though not trivial, I assure you. And that is, that for two-party
conversations the basic sequencing format is A-B Reduplicated. It’s not trivial
in that with three-party conversations it’s not the case that the sequencing
rules are A-B-C Reduplicated. There’s something else; what it is, I don’t
know. So: A-B Reduplicated. One party talks, then the other party talks, then
the first party talks again, etc. I use the term ‘two-party’ so as to provide for
the fact that this does not necessarily mean two persons. The ‘two-party’
conversation may be a basic format such that conversations having more than
two persons present can take a two-party form. That would involve persons
dividing themselves up into teams of a sort, and alternating according to team
membership, where, then, one team talks - a whole series of persons might
talk for that team - then the other team, etc.

Restricting our proper considerations to two-party conversation with the

The nine Winter 1965 lectures (all of them owed to ‘transcriber unknown’) pretty
much recapitulate the considerations of ‘Fall 1964,’ sometimes in a more developed,
formal way. Most of them have been incorporated into those earlier lectures: lecture
(1) has been absorbed into lecture 1; lectures (4) and (5) into lecture 6, lecture (6)
into lecture 9, much of lecture (7) into lecture 3, leaving lectures (2), (3), (8), and
(9). lectures (2) and (3) comprise this lecture 12, (8) and (9) comprise lecture 13.

95

96

Part I

sequencing rule A-B Reduplicated, what we have to come up with as a first
object is something we can call an ‘ adequate complete utterance.' And that will
be something that a person can say, which, upon its completion, provides for
the relevance of the sequencing rules. That is, on its completion, the other
talks, properly. A sentence may be complete, and one could tell that it’s
complete, but that wouldn’t tell you that the person is finished speaking for
now. But if they use an 'adequate complete utterance’, then, by virtue of the
fact that that unit is complete, the sequencing rules are relevant. Again, a
sentence is in general not sufficient, though some sentences may be as much
as a person is going to say. Nor, for example, is it generically the case that a
‘question’ is sufficient, though questions may comprise, and frequently do
comprise, complete utterances. That is to say, it is not enough to propose that
an item was ‘a question,’ to know that upon its completion the other was to
have talked, or did talk.

We want to see if we can get something that stands as an ‘adequate complete
utterance,’ such that upon the use of one or several of these, we have minimally
constituted something that will be, recognizably, ‘a conversation.’ This is not to
say that conversations are only built up out of adequate complete utterances,
because persons can have ways of detecting that something is or is not com-
plete, apart from the fact that the object is standardized in such a way. Though
insofar as they’re using other things than adequate complete utterances to
make up their talk, they have a special task of detecting that the other is or is
not finished, and perhaps what they ought to do now.

Another way that we could tell that ‘a conversation’ has taken place is if
there were some invariable part. We might then go about identifying the fact
that a conversation occurred by reference to the fact that the invariable part
occurred. Only, as far as I can tell, there aren’t any. However, there is
something pretty close to that. Take something like “Hello,” “Hello.” Now,
a ‘greeting’ is an ‘adequate complete utterance.’ It’s standardized as such.
When you hear ‘a greeting,’ then you can take it that when it’s complete, it’s
your chance to talk, if you’re the one that’s been greeted. There are several
things we want to notice about greetings, apart from the fact that they are
adequate complete utterances. Greetings are paired. And by that I mean
simply that if A picks a member of one of those things, then a proper move
for B when he has an opportunity to speak - right after it - is to pick a
member also; the same, or another. So one party’s use of a greeting provides
for that minimal exchange, “Hello,” “Hello.”

Now, it’s the case that if A-B Reduplicated is the format of conversations,
then there is no specific length that a conversation takes, to be ‘a conversation.’
And there may be no generic way built into the rules of conversational
sequencing, that a conversation comes to a close. So, for example, there can
be enormous variance between two conversations as to how much was said.
That is, you don’t have a situation where some certain amount of talk is
required before the conversation can, or ought to, close. Or, for example,
there can be an enormous variance as to how much one person has said, as
compared to the other. It’s not a situation where persons have to monitor how

Lecture 12 97

much they’ve talked as compared to how much the other has talked, to find
that the conversation can, or ought to, close.

Thus - if two things were so which are not so - we could say that we have
a ‘minimal conversation,’ “Hello,” “Hello.” And we could say that if at least
that took place, then a conversation occurred. And we could describe how that
could take place, given this A-B Reduplicated format, given an ‘adequate
complete utterance,’ given the ‘paired’ characteristic, plus a few minor things
which I’ll point to later on. But the two things are not so. One is, it’s not
invariably the case that things we would say are ‘conversations’ contain
greetings. The second is, it’s not invariably the case that ‘greeting items,’ such
as “Hello,” occur as ‘greetings.’

Now those facts lead us to require the following: We need to distinguish
between a ‘greeting item’ and a ‘greeting place.’ Where, then, something is
a ‘greeting’ only if it’s a ‘greeting item’ occurring in a ‘greeting place.’ If a
greeting item occurs elsewhere it’s not a greeting, and if some other item
occurs in a greeting place it’s not a greeting - though some items that are close
to greetings might take on the character of a greeting by occurring in a
greeting place. We need, then, to be able to say that there’s a ‘greeting place,’
and that any ‘conversation’ has it. And I take it we can say that there is a
greeting place in any conversation, by virtue of the following kind of
consideration.

First of all, it does seem that there is no rule of exclusion for greetings.
People can know each other 35 years, talk to each other every day, and
nonetheless greet each other when they begin a conversation. But take a whole
range of other items, for example ‘introductions’ (telling someone your name,
etc.). About introductions it can be said that there are rules for their historical
use. At some point in the history of persons’ conversations, introductions are
no longer relevant. And if they’re not relevant, then, when they don’t occur,
one can’t say they’re not there because there’s no reason to suppose that they
would be there. Notice that what we re trying to do is find some way of
saying, non-trivially, that something is ‘absent.’ If there were something that
was invariably present we would have no trouble. We could say if that thing
happens, then ‘conversation’ occurs. But we need to be able to say that we
have a conversation if that thing is present, or if it’s absent. And to say that
something is ‘absent’ is a much harder task. The way we can go about doing
that is to find that it’s always relevant. If it’s relevant, then if it’s not present,
we can say non-trivially that it’s not there. And greetings have that sort of
relevance, in that there is no rule of exclusion for them. So we can say that
greetings are relevant for any conversation.

Secondly, we can distinguish between the greeting item in the greeting
place, and the greeting item elsewhere. That is, somebody can say “Hello” in
the middle of a telephone conversation, where what they’re doing is not
‘greeting,’ but checking out whether the other person is still on the phone, and
a variety of other things like that. Thus, the greeting item, to be ‘a greeting,’
has to occur somewhere in particular, and we can say, then, that there’s ‘a
place.’ And in that greetings are relevant for any conversation, that place in

98

Part 1

which they are recognizably not something else, but ‘greetings,’ is present for
any conversation, whether there is a greeting item in it or not.

It’s not, then, that we just need “Hello,” “Hello” or members of that class
to have taken place, to have ‘a conversation’ and to warrant our being able to
say that there is a natural analytic unit, ‘conversation.’ But if we can say about
some piece of talk - either a greeting item or a greeting substitute such as
“How are you?” - that it occurred in ‘the greeting place,’ and that piece of
talk, whatever it was, provided for the relevance of the sequencing rules, then
we could say that we have ‘a conversation.’ And then we would be able to
warrant, at least in part, the fact that there is a unit, ‘conversation,’ which is
natural and analytic, and is generically usable. Of course the description of
cases of it may be far more complicated than this, but it has to be noted in the
first place that such a warrant has never been made, and that such a unit has
not been established. And it’s for that reason, at least, that it’s of interest.

Now, it does seem to be the case that “Hello,” “Hello” is a ‘minimal
conversation.’ Persons take it that it’s a minimal conversation. It’s not
sub-minimal; you don’t need more to have had ‘a conversation.’ And you do
need that, or substitutes for it. Children learn this at a rather young age, and
you find them producing perfectly recognizable ways of indicating that they
have engaged in the beginning of what may be only a minimal conversation,
and have not been properly treated. That is to say, one of the ways that one
shows that one has done something which is an adequate complete utterance
- that is, which is appropriate for the use of the sequencing rules - is to repeat
it. So I have these reports where a child says “Hi,” there’s no answer, and the
child says again, “Hi!” And then there’s a “Hi” in return, and the child will
take that as having been sufficient, and go about his business - which he
doesn’t do when he says “Hi” and there’s no return.

That use of repetition as a way of indicating in the first place that an item
was adequate for whatever it is that’s supposed to come next, is obviously the
simplest way of doing that task. In the child’s use of it, however, we get
something that’s worthy of some brief mention. And that is the way that
adults come to see that the child knows something of language. The way
adults know that the child is now ‘speaking’ and not babbling, involves the
fact that the minimally recognizable units of infant speech - and this is
essentially cross-culturally valid - are combinations of ‘p’ or ‘t’ or ‘d’ followed
by a vowel like ‘a’. And those combinations seem by and large to be used
without respect to what the language is; that is, without respect to however
the adult language may be constructed. And the way, apparently, that one
tells that the child is now speaking is by virtue of the fact that it doesn’t
simply produce a series of syllables, but it repeats a syllable. In this culture,
then, prototypically the first word that a child speaks is the word for ‘father,’
“da da .” 1 The interest of this phenomenon lies in the fact that if you get that

1 Sacks cites the paper in which “this is all discussed,” but the title wasn’t caught
by the transcriber. One possibility is Ruth Hirsh Weir, Language in the Crib (The
Hague: Mouton, 1962).

Lecture 12 99

kind of stability, across fantastically different languages, then the social
sciences and biological sciences come to some close relationship.

So, in the first place, this duplication business is a non-trivial fact, and it’s
pretty much as simple a way as you can have of indicating a range of things

- in this case, that something had been done, and was adequate for the
relevance of the sequencing rules. And you can notice the way that parents
point out to children their violations on this matter. Suppose somebody comes
to the house and says “Hi” to the child and the child doesn’t respond. One
thing the parents will say is, “Didn’t you hear them say ‘hi’ to you?” Where
they take it that they don’t have to restate the sequencing rules, but simply
point out that the rules have been adequately invoked.

1 take it we can say, then, that the unit ‘conversation’ is warranted by the
fact that we have at least a minimal thing that’s recognizable as ‘a
conversation.’ For it, the sequencing rules are relevant. We can talk about
places in it, or a place in it anyway, and by virtue of this we can also see that

- at least for the discipline of sociology if not for any lay interest in it - these
things, ‘greetings,’ are of some central theoretical importance, though
“Hello,” “Hello” looks like nothing that one would want to attend to very
much. Their consideration does an enormous amount of work for us. And
once we’re dealing with the fact that we’ve got sequencing here, and it’s
regulated, we’re no longer in a position where linguistic investigations are
usable. Because grammars don’t differentiate this way.

With “Hello,” “Hello” and things like it - members of a class of paired
activities such that if A uses one, B’s proper move is to use one also - we’re
examining the sequential building blocks of conversation which are specially
relevant in terms of their sequential character. Now, in that light, we can
consider jokes. The following comes from a group therapy session; the
members are teenagers. 2

A: Hey wait I’ve got I’ve got a joke. What’s black and white and

hides in caves?

B: Alright I give up. What’s black and white and hides in caves?

C: A newspaper

A: No. Pregnant nuns.

silence

B: Whyn’t you run across the street and get me some more coffee?

A: Why don’t you drop dead?

D: Whyn’t you just run across-?

A: What’s black and white thump black and white thump black

and white thump? nun rolling down stairs.

D: You know what a cute one is. You want to hear what a cute one is?
What’s purple and goes bam bam bam bam A four-door plum.

2 The fragment is taken from Sacks’ research notes and is slightly different from,
and closer to the actual data than, the one in the transcribed lecture.

100

Part I

A: Terrific

D\ I think it’s much better than about a black and white nun going
downstairs

A: No. that’s the new fad. Instead of having elephant jokes, now it’s nun

jokes.

B: Nothing. A nun.

D: hmmmmm

A: hehhhhh

C: What’s black white and grey? Sister Mary elephant.

A: hehhhhh

B: Say whata nuns really do? they must have some function?

C: r Nothing

A: 1 They travel in pairs.

B: Nothing nothing hehhh

A: They travel in pairs. One nun makes sure the other nun don’t get

none.

B: You know what’s a ball. Whistle at ’em whistle at ’em when they

walk down the street.

D: You know they usually ah pray.

A: Yeah.

D: That’s about all.

B: But the rest of the church does. Or they pray harder.

A: They’re women who have devoted their lives-

D: They marry God.

A: -to God

A: No. they’re women who have devoted their lives-

B: They’re women who’ve had a bad love life and become nuns.

A: -their lives their lives to uhm the devotion of the church.

B: J.C. and the boys.

D: We’re on an awfully bad God kick.

A: OK let’s change the subject.

The first thing that’s important about jokes is that to use one is something
like buying a drink among a bunch of people: They come in rounds. And if
some person tells a joke then every other person present has the right to tell
a joke. So we can say about a joke when it’s used, that it’s a ‘first joke’ and
that it will provide the occasion for each other person present to have a chance
to talk, and to have a particular kind of chance to talk; that is, a chance to tell
a joke.

So what? So there are a variety of cases where you get more then two
persons present, where exactly what the rights to talk are of the various
persons present, may be quite obscure. For example, there may be large status
differences or a variety of things like that, and how one goes about providing
that each person can talk under such circumstances may be, then, a real
question. For someone to use a joke on that occasion is then to give each other
a set place to talk, and also to give him something to say.

Lecture 12

101

And in this group therapy setting, those facts are quite non-trivial. The
persons are there for two hours, and that they keep talking is absolutely
crucial, for the occasions of silence are extremely dangerous to all persons
concerned. What happens when they’re silent is that various persons in the
place now begin to look for a face that’s noticeable among themselves, and
then pose for that person the problem of giving an account of why he is silent.
Or, for example, if any given person is silent for any length of time, then their
silence is a noticeable fact; something about which they can be questioned.
Now, if they want to raise their personal problems they’re quite free to do so,
but if they don’t want to, and given that silence is something noticeable, then
it’s important to have some sorts of things that will permit everybody to talk,
where they can talk without saying anything that can be ‘used against them.’
It also seems to be the case that any given body of talk, starting at any given
place, will, if allowed to go on, end up dangerous. That is, it will end up on
some topic which is perhaps too important to be talked about except under
real feelings of relaxation. Things like God, death, sex, for example, which
always come out, whatever topic is started. And persens are - and the people
in this therapy group are - much given to watching when it is that a topic
looks like it’s about to shift into something that from the group’s point of
view is to be avoided - though any given person can talk about it
‘themselves.’

What seems involved, then, is the development of things which permit
talk to go on, and to go on in an ‘unaffiliated’ manner. Notice about jokes,
that when jokes are told they’re things that are ‘going around;’ they’re quotes.
So they’re unaffiliated remarks, and in that sense it’s hard to say about
somebody that the fact that they told some particular joke has some special
significance. They just heard it, and now they’re repeating it.

Persons can then monitor the conversation, watching either for silence or
for the approach of something dangerous, and start a bloc of talk by flicking
in a joke, thereby giving each other person their chance to talk, and to talk
‘safely.’

I’ll just note here that there are other things which have the ‘unaffiliated’
character of jokes - that is, the speaker does not disclose his position by using
it - but which don’t have that sequential character of going in rounds, which
seem to get used in similar circumstances. At least these kids use them, and
until I noticed that they were being used in similar circumstance to the jokes, I
found them puzzling. What they do is, at points in the conversation when
either nobody is talking or they haven’t talked for a while, insert slogans.
They’ll just come out with a piece of an advertisement from the radio, or a
jingle, or obvious quotations sarcastically said. Again, then, it seems that they
go about monitoring when they ought to be talking or when silence seems to be
present, and flick out these things which, again, have this unaffiliated character.

Let me turn now to another sort of sequencing issue. I said about “Hello,”
“Hello” that it’s a paired phenomenon, and that when the second one comes
out, it may well be that the conversation is complete. And we can note that
either one, or both, can be delivered in such a way as to provide for the size

102

Part 1

of the conversation. For example, by not slowing down at all as you pass
somebody. You say “Hello,” the other says “Hello,” and there’s no indication
by either party that there’s going to be any more to it than that.

Now, some of the ways that conversations can begin, provide at least a
slightly different set of sequential characteristics, although there are ways in
which they’re related to paired beginnings. Perhaps the best way to introduce
it is by just reporting what I did the last time I was trying to introduce this
some material, so as to indicate right off that it wasn’t as trivial as it looks.
Before I presented the phenomenon, I’d asked people in the class to write
down the first lines of what they took to be ‘pickups.’ I got 60 first lines, of
which just under 60 were questions.

What I had wanted to be saying to them, and which they could see once
they had those collections, was that a person who asks a question has a right
to talk again after the question has been answered. So, with a question
beginning, the conversation goes at least something like A-B-A. It can go on
from there, or it can end like that. And that may be without regard to what
the question consists of or what the answer consists of. Now, one way that the
conversation can go on from there is that the person who asks the question can
use his initial right to talk again, to ask another question, and the same right
holds. So you can get indefinitely long chains, running, Q-A, Q-A, Q-A, etc.
Eventually I’ll go over the special relevance for certain conversations of the
‘chain’ possibility. It turns out to be extremely important. Whenever it
happens to occur in a conversation - and it doesn’t necessarily have to occur
in the beginning, but that point where somebody starts questioning - then the
‘chaining’ possibility can be quite crucial to the way that the conversation
goes. 2

Now, of the sorts of questions that occur in first conversations, let’s begin
by looking at those which have a close relationship to “Hello,” “Hello.” Note
that the use of “Hello” is a regulated matter. It is the sort of thing which can
be used to begin a conversation where two persons have some initial right to
talk to each other, such that the fact that they happen to be physically
co-present provides the occasion for the conversation. But, especially for
things like pickups, the fact that the two persons are physically co-present is
not sufficient grounds for them to begin talking, and “Hello” may be
inappropriate. You can get conversations which go:

A: Hello

2 This, and the following materials, constitute a next run at some of the
phenomena considered in lecture 7 of this Fall 1964-Spring 1965 series. These
materials were not incorporated into that earlier lecture although such incorporation
has been done with most of the Winter 1965 materials, because in this case it would
introduce an anachronism. Specifically, in this second run we see the first reference to
“the ‘chaining’ possibility,” which later crystallized as ‘the chaining rule’ (see Spring
1966, lecture 2, p. 256 and lecture 2 (R), p. 264). Not incorporating these materials
into the earlier lecture makes for some repetition, but the genesis of ‘the chaining rule’
seemed worth preserving.

Lecture 12

B: (No answer)

A: Don’t you remember me?

103

Where that involves proposing that there had been an initial right to use
"Hello.”

In the absence of some obvious warrant for the conversation to take place
by virtue of two persons being copresent and nothing else, you get that sort
of question which provides that although it doesn’t seem to be the case, there
is indeed a warrant. There’s a whole range of things which tend to formulate
a first conversation as a version of an nth. Things like "Don’t I know you
from somewhere?” “Didn’t I see you at such-and-such a place?” "Didn’t you
go to such-and-such a school?” “Aren’t you so-and-so?”, etc. All of these
provide for the fact that it may be the case that we know each other, and if
we do, then this conversation can take place as ‘a further conversation.’ In
those cases, then, and more generally, we have a class of questions which
provide an account for a conversation developing; that is, that this is not an
initial conversation.

There are a variety of other accounts which focus on different matters. For
example, very frequently the first question will be a request. And the request
will be such a thing as can be asked when any two persons are physically
available to each other; you’re standing in a crowd waiting for a plane to
arrive and someone asks, “When is the plane expected?” A variety of such
informational matters can be offered. Note that for any one of these, to
whatever the first answer is, another question can be constructed:

A: When does the plane arrive?

B: 7:15.

A: Are you going to San Francisco also?

It’s also to be noted that such standardized questions as "When does the
plane arrive?” "What time is it?” etc., by virtue of the fact that they are
standardized, provide for the relevance of the sequencing rules such that one
knows when one of those questions is complete. Further, one knows what an
answer to such a question looks like, so that the one who asked the question
can know when the thing that stands as an answer will have finished, and thus
provide that the other can talk again. This stands in contrast to discursive talk,
where it may not be clear in that fashion when it is that somebody has
finished. For persons who don’t know the discourse patterns of somebody
they’re dealing with, the use of standardized objects to build the beginning of
a conversation may be quite important. First of all, that you don’t wait too
long after the other has stopped, where waiting too long might provide for
their withdrawal altogether. Secondly, that you don’t interrupt, where of
course one doesn’t want to be ‘rude’ to someone with whom one is making
an effort to get acquainted.

((Thereafter is a discussion of the question "Do you have a light?” See
lecture 7, pp. 50-1.))

Lecture 13
On Proverbs

I’m going to talk about proverbs, trying to develop what’s interesting about
them. I’ll begin by doing something I don’t normally do, which is to read you
the way that proverbs are largely used by social scientists - because it’s quite
relevant to the task they seem to have set themselves. The first quotation
comes from page 3 of a book called The Study of Thinking by Jerome Bruner
and some associates of his.

That there is confusion remaining in the adult world about what
constitutes an identity class is testified to by such diverse proverbs as
‘plus ga change, plus la meme chose’ and the Hericlitan dictum that we
never enter the same river twice.

A very similar sort of remark comes from pages 1-2 (these always come in the
first several pages of a book) of George Homans’ Social Behavior: its
elementary forms.

My subject is a familiar chaos. Nothing is more familiar to men than
their ordinary, everyday social behavior; and should a sociologist make
any generalization about it, he runs the risk that his readers will find him
wrong at the first word and cut him off without a hearing. They have
been at home with the evidence since childhood and have every right to
an opinion. A physicist runs no such risk that the particles, whose social
behavior in the atom he describes, will talk back.

The sociologist’s only justification is that the subject, however
familiar, remains an intellectual chaos. Every man has thought about it,
and mankind through the centuries has embodied the more satisfactory
of the generalizations in proverbs and maxims about social behavior,
what it is and what it ought to be . . . What makes the subject of
everyday social behavior a chaos is that each of these maxims and
proverbs, while telling an important part of the truth, never tells it all,
and nobody tries to put them together . . . every man makes his own
generalizations about his own social experience, but uses them ad hoc
within the range of situations to which each applies, dropping them as
soon as their immediate relevance is at an end, and never asking how
they are related to one another. Every one has, of course, every excuse

A combination of Winter 1965, lecture (8), and lecture (9), pp. 1 and 8-12
(transcriber unknown on these two) and Fall 1964, tape 13, side 1.

104

Lecture 13

105

for this shortcoming, if it be one. Social experience is apt to come at us
too fast to leave us time to grasp it as a whole.

I don’t intend to make any detailed comments about those remarks, except
to note parenthetically that Homans’ procedure for starting a book is one of
the most recurrent you’ll find, and in a way it’s enough to tell you about the
kind of book you have here. He has to have an excuse to study the
phenomenon he wants to study. That it happens is not a sufficient excuse. He
has to show some problem. And he starts off with the supposition that persons
think they know about the thing he wants to study, so he finds a way to show
that they don’t. Now, to notice that is to notice that in the ordinary world, in
everyday life, ‘engaging in inquiries’ is an accountable thing. Where, then, the
work of sociologists remains constrained by that format.

I want to focus on this sort of thing, available in both these quotes: It’s a
very usual use of proverbs among academics, to refer to them as ‘propositions’
and to suppose then that it goes without saying that the corpus of proverbs
is subjectable to the same kind of treatment as, for example, is scientific
knowledge. They then build the basis for an inquiry - which has nothing to
do with proverbs - by virtue of the fact that these propositions, when
compared - without showing that they are actually compared in their use -
are inconsistent.

So one question is, are the collections of proverbs indeed a set of
propositions in the sense that Homans proposes? Do you find any reason to
suppose that that’s so? For example, here’s something from a newspaper:

Premier Krushchev’s removal was viewed in Paris today as a serious
blow to the improvement of relations between the Soviet Union and the
Western powers. The primary reaction of government officials and
diplomats was surprise. But they were also deeply concerned. “Better
the devil you know,” one diplomat said, “than the one you do not
know.”

Would somebody seriously say to this fellow, “What’s the evidence for that?”
If it were a proposition in the first place, then the statement “How do you
know?” which is used not only among scientists but is offered by Members,
might occur.

But one of the facts about proverbs is that they are ‘correct about
something.’ That fact is especially important since some of them contain rules,
and are invoked to govern various situations. Now, for many other kinds of
rules, even in highly rule-controlled situations like the legal courts, if you
invoke a rule by reference to precedent, the occasion of using it can provide
the occasion for reconsidering that rule to see whether, not only in this instance
but in general, it ought to obtain for anything. So that a rule introduced to
govern a situation in a law case can be changed altogether. Which is to say,
even a strict precedence system such as that, doesn’t have objects as powerful
and as limitedly attackable as proverbs.

106

Part I

It’s in part that this is so that makes it quite irrelevant when a proverb was
established - and in that sense, Bruner’s citing of “the Hericlitan dictum that
we never enter the same river twice,” which is several thousand years old, is
a correct use of proverbs - whereas it might be relevant in, say, a lawsuit,
when a rule was established and what the circumstances were that generated
it. So, for example, in current civil rights cases you occasionally find an
extremely old statute introduced, where, then, the fact that it was introduced
many years ago by reference to, say, the buying and selling of slaves, can be
used to argue that it really isn’t worth anything now. For proverbs, such
considerations are just quite irrelevant. Likewise, the organization or the
society within which they were established is often almost as irrelevant. Where
an American court would never think of using a Russian precedent, those
Russian proverbs we come to know are treated as quite appropriate in
ordinary discourse. For one, then, the object ‘proverb’ is enormously widely
found, and, further, many proverbs are applicable quite across cultural
boundaries.

Now, aside from the social scientists’ orientation to proverbs, anything I’ve
ever looked at on that matter involves a list of proverbs and where they come
from, their age, variations, etc. Nobody seems to deal with actual occasions of
their use. And that’s because it is the folklorists, with their particular interest
in proverbs, who have been collecting them. I want at least to make a start on
considering proverbs in terms of occasions of actual use.

There is a class of proverbs known as ‘proverbial phrases,’ one set of which
I’ll focus on; things like “You’re stacking the deck,” “He’s hitting below the
belt,” “You’re way out in left field,” etc. This set of proverbs comes from
domains which have clear parameters. In baseball there are demarcated areas,
so that “left field” is a locatable place within that domain. And “stacking the
deck” is a locatable violation in playing cards. It seems to me that, for one,
this sort of proverb may provide a clarification of the sense in which we might
talk about ‘families of actions.’ And by ‘families’ I mean this kind of thing:
Wittgenstein talks about ‘families of games’ and proposes, for example, that
there is some intersection of rules between games in a same ’family.’ I once
tried to see what such an intersection of rules might involve, with the
following kind of trick. We took violations of rules in one game, which were
not violations in another game, and began to use them just to see what might
happen. So, for example, if you’re playing cards, it’s not proper to stand
behind your opponent and look over his shoulder, but there’s no rule about
it in chess. It’s nothing. So we did it in chess, and it did cause some kind of
disturbances. There was a sense that somehow something must be wrong.
And that perhaps had to do with the fact that it was a violation of a rule in
some game. It seems, then, that there may be a sense in which rules in games
can carry over into situations for which they haven’t been specified as rules.

And for ‘families of actions’ we might talk about the following sort of
situation. For some kind of activity there may be an event which is, in that
activity, not regulated. It’s not even seen as an event; there’s no language for
it, or if it’s pointed out descriptively, nothing much would be said about it.

Lecture 13

107

But if you invoke a rule from another activity - especially those which can be
produced in proverbial form - then you find that you can come to see the
event as indeed an ‘event,’ and as possibly illegitimate. So, for example,
suppose we have a meeting, and there are various interests present at the
meeting. The way in which persons arrange their seating may not be treated
as something noticeable, and if pointed out would only be treated as, “So they
happen to be sitting in some order, so what?” But with respect to playing
cards, the way in which a deck is shuffled is something that’s regulated. And
there are clear ways of violating that, which are observable, and which are
talked of as “stacking the deck.” And when we propose at a meeting that
“the Commies are stacking the deck,” then it’s clear what it is that’s going on,
and it’s treated as something suspect.

So apparently one uses proverbs of this character to make events noticeable,
perhaps to make their ordered character noticeable, and then to formulate
their ordered character by reference to their possible illegitimacy - where there
is in the first instance no rule governing this or that particular event. We
might loosely talk of this as a matter of ‘analogy,’ though it only becomes
analogy once it’s proposed that some order is relevant in both circumstances.

What we can come to see is that there can be a very limited set of
paradigms or models, each of which may have demarcated areas of order
which can operate generatively for an enormous range of further areas -
without, however, it being necessary to produce a further set of rules at all, or
to further demarcate the parameters of this other mass of events. Once one
knows that “stacking the deck” is something that can be done with a set of
cards, it can be seen that something like it’ is being done with a set of people.
Or, for example, where “hitting below the belt” is something that is done in
an exchange of blows, it can be seen as comparable to something that can
happen in an exchange of remarks, and thus to operate in conversation and
other sorts of exchange activities. One doesn’t have to construct a new
language, but can retain the base source of this or that rule, as providing the
terms.

I want to turn now to a consideration of proverbs as pieces of the language.
In Archer Taylor’s classical book The Proverb, he mentions in passing one
aspect of these things, which got me started on the line I’ll be discussing.
Right at the beginning of the book he’s talking about defining proverbs:

The definition of a proverb is too difficult to repay the undertaking; and
should we fortunately combine in a single definition all the essential
elements and give each the proper emphasis, we should not even then
have a touchstone. An incommunicable quality tells us this sentence is
proverbial and that one is not. Hence no definition will enable us to
identify positively a sentence as proverbial.

When I read the book a couple of years ago, what was important in that
statement just never occurred to me. I’ll formulate it in what may sound a
very queer way: If we take it with the linguists and the grammarians that the

108

Part I

sentence is constructed according to some rules, and that it is a kind of unit,
then one question we might be led to ask is, what other ways, aside from
linguistically, can we talk of a sentence as ‘a unit? What can be done with it?
That is, what are the sorts of activities that persons can hope to accomplish
within a single sentence?

Such a question is obviously relevant to the analysis of social activities.
Leaving proverbs aside for the moment, we could say that there would be at
least a special kind of tension in a world where single sentences could do
exceptional work. For instance, in the phone calls I collected at an emergency
psychiatric hospital, you’d occasionally get the following. A man might report
that he’s suicidal and offer as the immediate warrant that after 25 years of
marriage his wife said, “Well, the fact is I never really loved you.” A
statement like that has a dual consequentialness. It not only can bring
something to a close, but it can provide that some series of events have been
falsely seen; that is, it can erase them.

And there is a whole set of things like that. Some religions are carried on
that feature. The fact that you can know that at any last moment - if you get
a last moment - you can always change your ways sufficiently to erase
whatever it is you’ve done in the past, means that you do not have to order
yours sins day by day. That is, no matter how your sins have added up, it’s
possible at any last moment to change, and to get what anybody can get no
matter what they’ve done earlier. Of course the fact that you may not get that
last moment, itself sets up a variety of other kinds of considerations. But the
sort of religious existentialism which focusses on every moment as a possible
last moment, is itself only relevant given the notion of what can be done in
any given moment.

So the issue of what can be packed into some single unit is obviously
interesting. And with respect to language, we can examine the variety of ways
that the resources of a language have been explored in terms of how much can
be packed into something while retaining such central features as, for
example, transmissibility and reproducibility. Now, poetry and proverbs have
taken a quite different tack from other explorations. It is said that poetry is
very similar to mathematics, in the sense that you pack in knowledge with far
more economy than prose would offer. And that is shown in elementary
mathematics books when they ask you to “rewrite in English the following
equation.” The same kind of test can be done for a poem. And it may be
recalled that early Greek science, for example, did use poetical forms to write
up the results.

Now, one of the crucial things about proverbs is that they’re objects from
an oral body of knowledge. They do indeed get written down, but their basic
power and relevance seem to be as oral objects. There is a literature on oral
traditions and how they’re preserved and used, and I’d like to suggest a most
extraordinary book on this subject which is misleadingly titled if you’re
looking for material on oral traditions, and that’s Preface to Plato by Eric
Havelock. It is about Plato, but you could have no interest at all in Plato and
learn an awful lot from it. Its basic concern is, what is Plato up to in his attack

Lecture 13

109

on Homer? Havelock argues that for the Greeks, Homer was an encyclopedia.
His poems stored the enormous amount of relevant knowledge that the
Greeks had to use, where the Greeks in this period did not, except in very
exceptional circumstances, use writing. Homer’s poems were one of a variety
of very powerful devices used to store that information. Plato’s concern was
to break down that way of preserving knowledge because of a variety of
things that bothered him about it. It’s not that Plato didn’t like poetry, in the
modern sense of poetry, but that he was aware of the ways in which poetry
is powerful - and the limits of that kind of power.

I couldn’t begin to give an elaborate discussion on the relation of poetry
and proverbs, but I can say a few things. Proverbs, like poetry, have a large
use of metaphor, and they often have a kind of alliteration, rhyming, etc.
There are more or less standard forms which are used, to which any given
instance is fitted as far as I know; proverbs are formed as single sentences or
phrases of sentences, they’re not longer than that. And, like poetry, their
reproduction consists of the exact repetition of them - a poem and a proverb
lose their character when they’re paraphrased. These sorts of features have
some real advantages for maintaining a body of knowledge; we can say that
they are constructed in mnemonically efficacious ways. That is to say, they’re
very easily remembered and are thereby transmissible as ‘that very item’ and
not in a paraphrase. Their stability, then, can be something independent from
any occasion of use.

Now for proverbs, I take it that one of the core features of their sense and
of their use is that they are ‘atopical’ phenomena. So, for example, the sense
and relevance of ‘a rolling stone gathers no moss’ is not found by reference to
geological or botanical considerations. Some of the work of the neuropsychol-
ogist Kurt Goldstein and his associates may be relevant here. One of the
things they’ve found for children, brain-damaged persons, and sometimes
among schizophrenics is that a kind of test devised by psychologists indicates
that these people cannot handle proverbs - they don’t understand them, they
don’t know how to use them. There are many protocols of persons presented
with a proverb and asked to interpret it, and they produce long discussions
about various features and behaviors of, for example, stones and moss.

Goldstein proposes on the basis of those tests and other indicators, that
there’s a big split between what he calls ‘abstract’ and what he calls ‘concrete’
thinking, and that persons who can’t use proverbs are persons involved in a
failure of abstract thinking. But I take it that if you look at the protocols, the
persons involved seem to be quite capable of dealing with proverbs
‘abstractly.’ First of all, the proverbs themselves are quite abstract. ‘A rolling
stone gathers no moss’ doesn’t contain any reference to a particular rolling
stone, a particular kind of moss, etc., etc. And these people who are not able
to deal with proverbs properly are nevertheless talking abstractly. They’ll talk
about “a stone” and how it might roll, and say this or that about moss,
without any insistence that it has to be some particular stone.

That seems to me to stand in contrast to ‘concrete’ thinking. So, let’s say,
when you talk about some kinds of schizophrenics being “enmeshed in the

110

Part l

concrete,” you might be pointing to certain kinds of strange things they can
do that nobody else can do. For example, one day I was sitting having lunch
with a friend of mine and she said, ‘‘What day is it today?” I told her what
day it was, and she said “Last year on this day I had the following for
lunch . . . two years ago this day I had the following for lunch ...” and she
went through ten years, just spieling out the details of her menus. That, I take
it, is a pretty clear example of concrete thinking. But the troubles in dealing
with proverbs were not of that sort. Again, if the proverb contained the item
‘cat,’ then people would talk about cats and use the plural term ‘cats.’ They
did not start talking about ‘my cat.’

This suggests that we may not be dealing with inability to do ‘abstract’
thinking, but an inability to do ‘atopical’ thinking. Where, then, proverbs can
be seen to constitute a very clear example of whole collections of pieces of
knowledge that are organized atopically. And I take it that it may be this
feature that I’m calling ‘atopical organization,’ that Homans proposes as a
possible ‘shortcoming,’ in the remarks I quoted earlier:

What makes the subject of everyday social behavior a chaos is that each
of these maxims and proverbs, while telling an important part of the
truth, never tells it all, and nobody tries to put them together . . . but
uses them ad hoc within the range of situations to which each applies,
dropping them as soon as their immediate relevance is at an end, and
never asking how they are related to one another.

But there are some obvious virtues to having a body of knowledge organized
in an atopical fashion. You get a piece of knowledge, like ‘a rolling stone
gathers no moss,’ which is in the first instance correct about something. If you
paraphrase it into some particular domain, like “a man who doesn’t settle
down doesn’t gather possessions,” then it may not have the same kind of
correctness; it may be questionable. And as I mentioned earlier, one of the
most striking things about proverbs is that while on any occasion of use they
may be, for example, inappropriate, people do not propose about them, “Is
that so?” “What is your evidence for that?” If there is a question about them,
it is in terms of, is it appropriate to apply that proverb to this person, activity,
etc.? That is, proverbs are in the first place correct. And that can be
accomplished by formulating a proverb from a domain within which it is
correct, and having it always be used elsewhere. In that way, instead of
constantly revising a body of knowledge by reference to the discovery that it’s
not correct here, now, for this, you maintain a stable body of knowledge and
control the domain of its use.

I’ve already mentioned the feature of ‘single-sentence packing’ for prov-
erbs. Let me offer a few more remarks. It’s the case that both maxims and
descriptions can be produced as proverbs, formulable as single sentences. “A
woman’s place is in the home” is an instance of the former, and for the latter,
they are frequently those ‘proverbial phrases,’ “stacking the deck,” “hanging
by a thread,” “barking up the wrong tree,” etc. Both types of knowledge,

Lecture 13

111

then, can be had and used via the single sentence, and thus it’s not necessary
to have some combination of sentences so as to minimally understand and
transmit such information. Having that, you have a setup designed to permit
you to learn new members of that class of information much more quickly
than might be otherwise possible. That is to say, at that point when children
can make single sentences, their task is now to see what it is that can be
packed into that form they’ve learned. And that provides that they will have
solved the theoretical problems of being able to know norms, or know
knowledge. They would not then be faced with such a task as learning which
set of sentences can constitute a piece of knowledge. Rather, once the single
sentence is gotten, a whole series of things can be fitted to it, instead of having
to deal with various combinations, which might set up quite different sorts of
tasks.

In this regard, an extremely important thing about proverbs is that they
have the character of being potentially descriptive or relevant. Persons learn
them and have them available for use. They don’t, that is, learn them on the
occasion of their appropriate use. We could imagine that humans were built
such that their language was alike to that of other animals, and to the way in
which young children very heavily use language; either as narrative comment
on what they’re doing, “Now I’m putting on my shoe,” or only uttering
something on the occasion that it’s appropriate - though it may be wrong,
they could say “That’s mommy” and it’s not mommy; that is, the domain of
application would be correct or incorrect, but the thing is not uttered apart
from some possibly appropriate occasion. But you can have these potential
descriptions and see them as ‘correct for something,’ where what it would be
correct for remains to be seen.

And if you watch kids learning language, then learning that sort of thing
seems to be an important part of it. Observations have been made of children
talking to themselves before they fall asleep. And it’s been found that they’re
not coming out with a bunch of random noises, but with a training procedure.
For example, they’ve been observed doing exercises in phonetics, combining
and assembling phonemes. I don’t think it’s been pointed out that once they
have words, they play with combinations of terms that involve building
possible descriptions. We could say it would be strictly a matter of learning
the linguistic features of the language - learning the language in this very
technical sense - that kids begin to use, in close relation to each other, the
paired antonyms (things like in-out, up-down, etc.); that is, to produce these
paired items as pairs. However, if they not only did that, but did - as I think
they do - use them in proper sequence, then we’ve got them learning
something more. They’re learning potential descriptions. That is, they don’t
say, for example, “He fixed it, he broke it,” but “He broke it, he fixed it.”
And for such things as in-out, which in some cases have proper sequences and
you don’t turn them around, then they do them in proper sequence, “Kitty
got into the box, he got out.” And if they do turn them around, they do it
properly: “Kitty got out of the box, he got in again.” They’re learning, then,
not only the relation of certain types of words, but that strings of words can

112

Part I

assemble potential descriptions, only in certain arrangements. And that is to
say, they’re learning that there is a correct way of assembling potential
descriptions, apart from the particular occasions of their use.

Now, the phenomenon of ‘potential description’ has a variety of special
consequences. One of them is that it sets up the possibility of a logic which
involves examining statements apart from an occasion of use - and by ‘an
occasion of use’ I mean an application to some actual situation, not, for
example, an attempt to make a proof. Another is something which I had to
be brought to see as noticeable, and that is the possibility of literature; stuff
that is about nothing in particular, in which an author isn’t talking about any
actual set of events. It occurred to me that the phenomenon of ‘the possibility
of literature’ is something noticeable while I was looking at psychiatric reports
about delusions that patients have. In these reports, the writers take it that the
delusions are understandable, though what makes them ‘delusions’ is that
they couldn’t possibly describe something. Literature has a similar character;
it’s composed of possible potential descriptions and possible potential rules,
and a reader can look at assembled strings of language and decide that it’s
‘realistic’ or ‘not realistic,’ compelling or not compelling. And in that sense,
the possibility of literature and the possibility of logic are very, very closely
related.

These sorts of considerations may have a bearing on a classical controversy,
mainly within linguistics, which concerns the question: Are grammar and
meaning separable? It may be that “Are they separable?” is not the problem,
but that if ‘meaning’ is, in part, reference to something, and grammar is
understandable apart from reference to anything - that is, formally correct in
some way - then it’s not simply a technological linguistic question, whether
it’s so that they’re separable, but that it’s an essential fact of language that
they are separate.

Lecture 14

The Inference-Making Machine

(1) A: Yeah, then what happened?

(2) B: Okay, in the meantime she {wife of B} says, “Don’t ask the child

nothing.” Well, she stepped between me and the child, and I got
up to walk out the door. When she stepped between me and the
child, I went to move her out of the way. And then about that time
her sister had called the police. I don’t know how she . . . what she

(3) A: Didn’t you smack her one?

(4) B: No.

(5) A: You’re not telling me the story, Mr B.

(6) B: Well, you see when you say smack you mean hit.

(7) A: Yeah, you shoved her. Is that it?

(8) B: Yeah, I shoved her.

One of the basic things I want to be able to give you is an aesthetic for social
life. By that I mean in part that we should have some sense of where it is deep,
and be able to see, and to pose, problems. I’ll try to do somewhat more than
that. I’ll also try to develop a variety of notions of what kind of business
sociology is, what its problems look like, what the form of the solutions to
those problems are, and perhaps to some extent, some of those solutions.

The kind of phenomena we are dealing with are always transcriptions of
actual occurrences, in their actual sequence. And I take it our business is to try
to construct the machinery that would produce those occurrences. That is, we
find and name some objects, and find and name some rules for using those
objects, where the rules for using those objects will produce those objects. And
we also consider conversation per se, looking at the rules for sequencing in
conversation.

The quotation I started off with comes at about the fifth interchange into
a first telephone conversation, where A is a staff member of a social agency
that B has called. B was told to call this agency because of some marital
troubles he’s having. A doesn’t know anything about B’s marital troubles,
except what B tells him. So apart from the four or five previous interchanges,

A combination of Spring 1965, lectures 1 and 2(transcriber unknown), with
materials from Fall 1964, tape 12, side 2, pp. 1-6. Only three of the Spring 1965
lectures are extant. Lecture 3 does not appear in this volume. The large part of its
materials are covered in lecture 1, and a bit of it has been incorporated into that
lecture.

113

114

Part I

these persons do not know each other, they’ve never met. And - though it’s
not exactly correct to say it this way - we find that “nonetheless,” given
essentially B’s statement number (2), A is able to have some notion of what
it is that has happened, in a story that he has only heard part of. Without
knowing B at all, hearing “When she stepped between me and the child, I
went to move her out of the way. And then about that time her sister had
called the police,” A can say, “Didn’t you smack her one?”

And there is more to it than that. Because it would be one thing if A had
some expectation about what the sequences of events were that brought on
the coming of the police, where A would use that expectation to make a
guess, but if it turns out that he’s told it’s not so, then so far as he knows, it’s
not so. But apparently A has a stronger grasp of the situation already, in the
sense that it’s not merely that he can have a guess, but that he takes it
that - perhaps without regard to what B tells him - his guess is correct.

And we can notice that the fact that A has a guess here is not anything very
extraordinary for B. He doesn’t say, as a person sometimes says, “What
makes you say that?” He says “No.” And when it’s proposed that his “No”
is not correct, he doesn’t say, “Look, I’m telling the story. How the hell would
you know?” (Parenthetically, when I say “He doesn’t say . . .,” that is a
rhetorical device on my part for loosening things up. I don’t mean it in a
serious sense. I am not making a statement that is intendedly descriptive. To
propose seriously, descriptively, ‘B doesn’t say X’ is another order of
proposition.) But B takes it that A does know, and B corrects himself.

Furthermore, A is able to see as well, that B “isn’t telling the story.” And
I want to briefly focus on that kind of thing: Seeing lies. I want mainly to
focus on it because it seems that adult intuition may be misleading in trying
to see what the problem is that seeing lies poses. In one of the early classics
of psychoanalysis, a paper entitled ‘On the origin of the influencing machine
in schizophrenia,’ which was published in English in the Psychoanalytic
Quarterly, volume 2, in 1933, Victor Tausk reports on one of his patients.
The patient was a young schizophrenic girl, and one of her symptoms he
found in the following way. He often questioned her, and one time when he
was questioning her she started laughing at him. He asked, “What are you
laughing at?” She said, “Why do you always ask me questions? After all, you
know what I’m thinking.”

And Tausk worked at this symptom, that schizophrenics think other
persons know their thoughts. The problem had been posed: How is it that
schizophrenics come to think that others know their thoughts? And he tries to
solve this problem. Now, Freud’s comments upon the presentation of the
paper are included in the journal publication. He says, “That’s not the
problem at all. After all, when you learn at least your first language, you learn
it from your parents, from adults. And children must take it that adults,
giving them the concepts, know how they’re being used; know how the child
is using them. So the problem is not how is it that people come to think that
others know their thoughts, but how is it that people come to think so deeply
that others don’t know their thoughts?” Then, in a characteristic type of

Lecture 14

115

observation, Freud says that the crucial event is the first successful lie. That
event must be traumatic. The kid must have to say, “My God, they don’t
know what’s going on!”

Whether that’s so or not, I certainly don’t know. But intuitively I take it
that most adults don’t see things that way, and that it’s certainly not an
inconceivable formulation. And the interchange we re looking at has a form
which is very characteristic between adults and children; that is, the mother
says to the child, “What were you doing?” the child gives some answer, the
mother - who wasn’t there - says, “No you weren’t,” and the child then cor-
rects itself. And again, it isn’t the case that persons proposing that one is lying,
or “not telling the story;” are treated as doing something quite extraordinary.

Now let’s try to begin to consider what our task is if we’re going to build
a machine that could in the first place produce this conclusion, “Didn’t you
smack her one?,” with this piece of information, “When she stepped between
me and the child, I went to move her out of the way. And then about that
time her sister had called the police.”

A first rule of procedure in doing analysis, a rule that you absolutely must
use or you can’t do the work, is this: In setting up what it is that seems to have
happened, preparatory to solving the problem, do not let your notion of what
could conceivably happen decide for you what must have happened. So, for
example, when we get this kind of conclusion drawn at statement (3), you
might say to yourself, how in the world could anybody think so fast? Because
this just comes off. There is no pause more than three seconds between (2) and
(3). You’re going to say, “How? People aren’t that smart.” And therefore
what happened must be something very simple; something which will require
only a simple solution.

And that leads to our second rule. There is no necessary fit between the
complexity or simplicity of the apparatus you need to construct some object,
and the face-value complexity or simplicity of the object. These are things
which you have to come to terms with, given the fact that this has indeed
occurred. And insofar as people are doing lay affairs, they walk around with
the notion that if somebody does something pretty simply, pretty quickly, or
pretty routinely, then it must not be much of a problem to explain what
they’ve done. There is no reason to suppose that is so. I’ll give an analogical
observation. In a recent review of a book attempting to describe the
production of sentences in the English language - a grammar, in short - the
reviewer observes that the grammar, though it’s not bad, is not terribly
successful, and it remains a fact that those sentences which any six-year old is
able to produce routinely, have not yet been adequately described by some
persons who are obviously enormously brilliant scientists. Of course the
activities that molecules are able to engage in quickly, routinely, have not been
described by enormously brilliant scientists. So don’t worry about the brains
that these persons couldn’t have but which the objects seem to require. Our
task is, in this sense, to build their brains.

Now, what features do we need to build into this machine? the first thing
of course, is that it is an ‘inference-making machine.’ That is to say, it can deal

116

Part I

with and categorize and make statements about an event it has not seen. And
the first thing about the sort of events it can handle is that they can be
sequential events. In the interchange we’re looking at, if we keep things very
simple what we see is that if we have (a), (b) . . . (d), then if we suppose that
A is a user of the machine we’re going to have to build, he can find what (c)
is. He is not merely in a position to make a guess about (c), but there are
stronger features to this machine. As I said earlier, it’s not that he guesses
something, is told “No” and says “Okay.” For him, there is some fit
operating between (a), (b), (c), and (d) such that the fact that he’s told that
what he proposes as (c) has not occurred does not constrain what he can
continue to insist on; that it was that (c) which occurred. And you can just
quickly contrast that to the situation of a riddle. I ask you a riddle and invite
you to give an answer. You give an answer and I say “No, the answer is . . .”
something else. And by definition that’s the answer. Where people do riddles
you don’t get an insistence, “No, I’m right.” The inference-making machine
we are building can handle riddles, but riddles set up a simpler task than
those which this machine is capable of handling.

Now let’s begin to examine, in a rather informal way, how it is that the
machine-user, A, seems to make conclusion (3), “Didn’t you smack her one?”
and then (5), “You’re not telling me the story, Mr B.” And here we can use
that information which we have as members of the same society that these
two people are in. What we have is roughly something like this: A knows that
the scene is ‘a family problem.’ So (a) is the family quarrel, (b) is the guy
moving to the door . . . (d) is the police coming. And (c) is the grounds for
the police to have come. That is, apparently on some piece of information the
police have come, and that piece of information is the thing that A has
guessed at. A apparently knows, then, what good grounds are for the police
to be called to a scene. And he’s able to use those good grounds, first to make
a guess, and then to assess the correctness of the answer to that guess.

We can note as well that it’s not simply the case that A and B don’t know
each other, but we have a set of other persons who are being talked about, and
A is listening to this. Whatever A knows about B, he certainly knows less
about these others. A knows essentially only the set of terms that B uses to
name them; that is, that there is something called a sister, something called
a wife, something called a child. It seems to me that the information that is
being used by A is held in terms of collections of these categories. For one, you
can easily enough come to see that for any population of persons present there
are available alternative sets of categories that can be used on them. That then
poses for us an utterly central task in our descriptions; to have some way of
providing which set of categories operate in some scene - in the reporting of
that scene or in its treatment as it is occurring.

To get a sense of the way in which the inferences that can be made from
a story are geared to these categories, we could try, for example, using
different categories. What if it were, not “her sister,” but ‘a neighbor’ who
had called the police? A possible inference in that case would be that the
grounds for calling the police had something to do with ‘creating a

Lecture 14

117

disturbance;’ crying child, husband and wife yelling at each other. Or, for
example, just shuffle the one category around a little bit. Would the same
inference be made if it was ‘my’ sister, not “her” sister who had called the
police? The rules with respect to who owes what to whom, and who takes care
of whom may be so formulated that those things matter a great deal. The
inference in this case might then be, not that the husband had produced some
activity which served as good grounds for calling the police, but that the wife
had done so.

And that is extremely important because it is an awesome machine if one
needs to know only that it is “my wife” and “her sister.” And you can do this
because that holds for every like unit in society, such that you don’t need to
ask for example, “Well tell me some more about your wife’s sister, is she
elderly? Is she prone to hysterics?” which is something that would be
absolutely essential in psychology. But if something like what I’ve been saying
is so, then: It is not merely that the notion that you need to know a great deal
about somebody before you can say this or that about them may be a lot of
nonsense, but the way that society goes about building people makes a
nonsense of such a notion. That is: A task of socialization is to produce
somebody who so behaves that those categories are enough to know
something about him.

One of the things we always want to be watching for is to see how simple
or how complex this animal is. In this regard I’ll raise a question but not
attempt to answer it here. There is what we can call an ‘order of depth’ in
dealing with various kinds of occurrences. So, for example, we can say about
A here that there can be a list of good grounds for police to be called to a
scene, and that list is built into his brain as a mature member of the society,
and he can, when told of some scene where the police come, now throw out
an item from that list as a guess, and perhaps furthermore insist on it being
correct. That would be one order of depth. But for the issue of how simple
or complex this animal is, we can notice that, for example, for dealing with
such a situation as ‘the police arriving on the scene,’ we need a machine; a set
of rules, that is, and a set of objects that those rules handle. We can then ask,
is that machine altogether distinct, such that if something else happens, we
need another machine for making another kind of inference, etc., etc.? Or will
it be the case that this use of an inference from the ‘police’ situation will be
absolutely similar to the way in which some other event is found? Then
police’ just becomes one category of a general machine that handles a whole
set of things. If that is so, then the task of describing how this animal operates
will be tremendously simplified. And the work of the animal itself is, of
course, tremendously simplified.

What we’re working with, for now, is that apparently A knows what the
good grounds are for the police to come, and uses those good grounds to
monitor what B tells him, and makes a guess and then challenges B’s assertion
that he’s wrong, I want show how this stuff can cut; how it can be interesting.
For example, it’s clearly the case - A knows it and B knows it - that the
police don’t only come for the good grounds that police come for. Cranks call

118

Part I

up the police all the time. And police answer calls that turn out not to be
based on good grounds. But apparently that fact has a special status. The
proper ways that the police come to some house are available to Members in
general, and it is those which may be used, for example, to decide what
happened at that house. And if you see somebody being led away by the
police, you may quite naturally feel that you know what’s up with him, at
least generically; that is, that he’s ‘done something wrong.’

Now that suggests two things: First, the fact that some procedure which
has a correct way of getting done, gets done correctly - independent of any
issue of police answering calls that are not based on good grounds, or any issue
of people getting harassed or misunderstood — may be quite crucial in
permitting persons to find the sense of an event which happened by reference
to those procedures. And second - and here we’re going to pose one of the
central dilemmas of Western civilization - a person who stands in the position
of having some procedure which has a correct basis for use applied to them,
stands in the position of having that procedure presumptively correctly
applied. And we can give that problem a name: Job’s Problem.

Job, the rich, good man, had lost all his wealth, his children, all his
possessions. His friends come to him, and there are series of long discussions.
What his friends propose is, look, you take it that God punishes the wicked
and rewards the good. We take it the same way. Your situation is
understandable only if you’re guilty. So confess. And for Job the question is,
“I don’t know that I’m guilty. I’m convinced that I’m not guilty. But then
how could this have happened to me?”

A central dilemma, then, is that some procedure which has a proper way
of operating, may not so operate. Kafka is dealing with the same kind of
issue. And Mr B here is in the same boat. What A suspects may indeed have
happened. Or B may have just felt . . . what can he say? After all, A knows
how those things happen, and how is B going to insist that it’s not so? That
is, a problem that people in a range of circumstances can be faced with is: Is
there a way available to provide from some event to have happened, apart
from the normal and proper way these things happen? We are talking about
something quite general here. Talking about it by reference to the police
doesn’t mean we are talking about a procedure set up by and for the police.
Rather, it may be the case that in order for the police to operate successfully,
they have to be able to produce their activities such that these ways that
Members have of looking at activities can be applied to the police. Likewise,
when I talk about Job, I’m not trying to deal with literary criticism. Job’s
problem focusses on a central problem in the way that persons go about
orienting to the occurrence of events, and that is, that it is somehow extremely
important that the inferences they do make can be taken as correct, and
thereby that those persons who produce those activities which are described by
these sequences so behave as to provide for the fact that these sequences do
describe them.

In that regard it is interesting to note that the phenomenon of presump-
tively correct descriptions, and behavior produced to fit those descriptions, can

Lecture 14

119

be found by reference to illegitimate as well as legitimate activities. So, for
example, in his book Order and Rebellion in Tribal Africa, in the chapter ‘The
reasonable man in Barotse law,’ Gluckman offers us the ‘reasonable wrong-
doer.’

The last case suggests that the Barotse have a picture not only of
reasonable and customary right ways of behaving, but also a picture of
the reasonable wrongdoer - the reasonable thief, adulterer, slanderer,
and so forth.

By this paradox - the reasonable wrongdoer - I sum up the fact that
wrongdoers in any society also behave in customary ways which are
socially stereotyped. There is the ‘criminal slouch’ as against the
‘scholarly stoop,’ the spiv’s clothes and hairstyle, the whole manner of
loitering with intent to commit a criminal action.

When there is only circumstantial evidence, these sorts of actions
build up before the judges until they conclude that . . . the total picture
is that of a reasonable wrongdoer; as we say, a person guilty beyond
reasonable doubt.

And I have heard cases which indicate that these customary ways of
doing wrong in fact influence adulterers and thieves, so that they give
themselves away in circumstances in which they could have acted so as
to cover up their misdeeds.

Our task, then, is not simply to be building a machine which can make
inferences, and make them in the strong way that I’ve proposed. A problem
for a sociology interested in describing socialization will consist in large part
of how it is that a human gets built who will produce his activities such that
they’re graspable in this way. That is to say, how it is that he’ll behave such
that these sequencing machines can be used to find out what he’s up to. Under
that notion, then, we would propose that at least one core focus for trouble
would be persons who are so socialized that they don’t permit these
sequencing machines to be used on them. And that is one way that
‘psychopathic personalities’ behave. If you read Cleckley’s book The Mask of
Sanity, the psychopathic personality is reported to be that person who, at any
given point in their behavior, you never know what’s going to happen next.
You’re never able to say “Here is an nth point in this sequence, and now X,
Y, and Z will come.” And they are taken to be about as painful a person as
you can have around you.

Now, what I have been proposing could be restated as follows: For
Members, activities are observables. They see activities. They see persons
doing intimacy, they see persons lying, etc. (It has been wrongly proposed that
people do not see, for example, ‘my mother,’ but what they ‘really see’ is
light, dark, shadows, an object in the distance, etc.) And that poses for us the
task of being behaviorists in this sense: Finding how it is that people can
produce sets of actions that provide that others can see such things.

120

Part I

Earlier I observed that one way that persons go about seeing activities is by
reference to some procedures which they take it properly occur as the activities
occur. That is, it can be seen that Mr B “smacked” his wife by virtue of the
fact that the police came, by virtue of the fact that the police come when
people smack their wives. While in many of its aspects the use of procedures
to find the sense of some set of observables - that is, the phenomenon I’m
calling Job’s Problem - is utterly central, this fact of observability of actions
is much more generic.

We can be led, then, to investigate, for example, how it is that persons
learn that by virtue of their appearances, the activities they have gone through
are observable. Where, again, this observability is not specific to each activity,
but is learned as a general phenomenon. And we could suppose that the
following sort of report might be the sort of thing we could use to find out
how this learning takes place, and to see where those things may or may not
be difficult. I will take this utterly mundane report and suggest some of its
relevance. It is a quotation from One Boy’s Day by Barker and Wright (195 1).
What they did was to have a bunch of people follow a kid around all day,
writing down as best they could, everything that he did. They worked in
half-hour shifts, and they compiled, then, a record of his day.

7:20 Raymond got up from his chair. He went directly out of the
kitchen and into the bathroom.

Coming from the bathroom, he returned to the kitchen. His
mother asked pleasantly, “Did you wash your teeth?” Mr Birch
looked at him and laughed saying, “My gosh, son, you have
tooth powder all over your cheeks.” Then both parents laughed
heartily.

Raymond turned instantly and went straight to the bathroom.

He smiled as though he were not upset by his parents’ comments.

He stayed in the bathroom just a few seconds.

He came back rubbing his face with his hands. The tooth
powder was no longer visible.

Presumably Raymond can learn through things like this, that his parents
can tell that he washed his teeth by virtue of the appearance on his face of
tooth powder. And that fact sets up the phenomeon which I’ll call generically,
‘subversion.’ With this example in hand one can think quite rapidly of the
way that children can learn subversion, having learned that the procedure is
applied to them. For example, they will wash those aspects of their body
which are at a glance observable, so when they appear, it seems “Oh, you
washed.” And the fact that that adaptation goes on, provides, then, the sense
in which they orient to this way that their activities are grasped.

Now, in watching these obviously trivial things that children might do, it
is conceivable that one may have analyzed matters that are treated as being of
larger moment. So, for example, the quotation I just gave, in which
Raymond’s having brushed his teeth was seen in his face covered with some

Lecture 14

121

white powder, is in its character extremely central, and deserves the name
‘generic.’ Simply enough, the first human event in Judaic-Christian mythol-
ogy consists of man’s discovery that his moral character is observable. We
might call it Adam’s Problem.

(Genesis, 3.6-12, King James Version)

And when the woman saw that the tree was good for food, and that it
was pleasant to the eyes, and a tree to be desired to make wise, she took
of the fruit thereof, and did eat, and gave also unto her husband with
her; and he did eat. And the eyes of them both were opened, and they
knew that they were naked; and they sewed fig leaves together, and
made themselves aprons.

And they heard the voice of the Lord God walking in the garden in
the cool of the day; and Adam and his wife hid themselves from the
presence of the Lord God amongst the trees of the garden.

And the Lord God called unto Adam, and said unto him, Where art
thou?

And he said, I heard Thy voice in the garden, and I was afraid,
because I was naked; and I hid myself.

And He said, who told thee that thou was naked? Hast thou eaten of
the tree whereof I commanded thee that thou should not eat?

And the man said, the woman whom Thou gavest to be with me, she
gave me of the tree, and I did eat.

And Adam learns, like Raymond learns, that one has to come to terms with
the fact that from one’s appearances the activities that one has engaged in are
observable.

Let me mention one further thing, which is again relevant to the
conversation we’re examining. The matter I’ll be dealing with is one which
you might not, given the way one goes about considering scientific materials,
formulate in the way I’m going to. When we think about facts, insofar as we
are thinking of scientific facts, we tend to pose problems in the following way:
If it’s the case that something has occurred, then our problem is to explain it.
Now, with such things as lies, untruths, confabulations - the possibility of
which persons are often attending to - we’ve got to notice that something like
a reverse procedure is very much used. The reverse procedure consists of the
following. In deciding among possible competing facts, one may decide that
that fact occurred which has an explanation, and that fact that hasn’t an
explanation did not occur. Here is a beautiful instance of this way of
proceeding. It comes from an arbitration case involving a company which
discharged some of its employees for causing a disturbance. Here’s what the
arbitrator writes:

(Shulman Decision A-70 Ford-UAW Arbitration)

122

Part 1

The story of the other discharged employees approaches the bizarre.
Twelve of them testified before me. Each of them claims to be a
completely innocent bystander, wholely at a loss to understand why he
was picked up for discharge. None of them admits being part of the
crowd in any of the demonstrations. None of them admits even the
normal curiosity of an innocent bystander. Each claims that he knew
very little about the cause of the stoppages, and cared even less after
learning the cause. Each claims that when the lights went out or the line
stopped he asked his foreman what to do and on being told to stay or
go home as he pleased but that his time stopped in any event, he left for
home. One of the men, a lively young boxer, asserts that after seeing the
crowd and the excitement he calmly repaired to a warm, comfortable
spot and went to sleep. He did this, he asserts, on two of the three days
(being absent on the third), and slept the peaceful sleep of the just, until
the excitement completely quieted down. All, it seems were veritable
angels, above and beyond contagion by the excitement in the depart-
ment.

Now there unquestionably were serious stoppages in Department 84
on November 5th, 6th, and 8th. There were vociferous and angry men
milling around and demanding action. Who were the incensed men
who did take part? Who were the angry men who were so difficult to get
back to work and were so incensed that, the union claims, they turned
against their own committeemen and even assaulted two of them? How,
indeed, were these fourteen chosen?

The union advanced no explanation. There is no suggestion that these
men were chosen by lot. And even such a method would normally be
expected to catch some of the guilty. And there is no basis whatever in
the evidence to suppose that the men were selected because of any
personal animosity against them, with the slight possible exception of
one man. Nor are the men generally regarded as troublemakers, of
whom the company would be glad to be rid.

The company’s explanation is simple and without any contradiction
other than the incredible stories related by the men themselves. The
labor relations conciliator, with the help of his assistant, took the names
or the badge numbers of the most active men in the crowds that
demonstrated in his office. This accounts for twelve of the fourteen.

. . . Under these circumstances I cannot give credence to the men’s
protestations of innocence.

In this case the two competing facts are that a group of employees are
innocent, were wrongfully discharged, and that they are guilty, discharged on
good grounds. If they are innocent, there is “no explanation” for their being
selected to be fired. If they are guilty, then the “explanation is simple and
without any contradiction.” The finding is that they are guilty. Here is
another instance of that sort of procedure, taken from a coroner’s report. The
competing facts in this case are death by suicide or accident.

Lecture 14

123

While Mrs S.’s drinking during the past twenty years may be
symptomatic of problems, there was nothing in the history to indicate
any sudden change in her life pattern, or any unfortunate or untimely
occurrence in her life. And therefore there seems to be no reason why she
should have chosen this particular time to end her life.

That is to say, since there is no reason for her to have ended her life, she didn’t
end her life.

This gets very, very subtle and curious at times, where you find the
following kind of situation. One recurrently problematic kind of death
involves alcohol and pills in some combination, where alcohol and pills in
combination are extremely deadly. That is to say, given some amount of
alcohol, far less pills are necessary to kill somebody than without alcohol. And
where persons die of alcohol and pills, with a few pills, then you get
something like the following argument proposed: It’s not suicide, it’s an
accident, because if they wanted to kill themselves they would have taken all
the pills they had, and they didn’t take them all. To which you might say,
well, but they did die, and maybe what they did was to kill themselves in a
perfectly efficient way.

One of the nicest cases I have is from the autobiography of an ex-mental
patient. It’s a very long description so I’ll give it in condensed form. He’s just
committed himself, he’s been there maybe a day or two and now he wants to
inform his family, but nobody will let him out of the ward to make a phone
call. He finally asks a doctor, who tells him “Wait here five minutes, I’ll look
for your file. If I find it you can be let out to make a phone call.” So he waits.
He stands there waiting for five hours. At some point a nurse comes by and
tells him to move away because he’s blocking the door. He says, “No, I have
instructions to stay here.” She says “From who?” He says “From the doctor.”
The nurse goes off and gets the head nurse, who asks him “Why do you
refuse to move from that spot?” He says “Because the doctor told me to wait
here.” The head nurse, as he notes, “couldn’t assume the risk of overruling
the doctor” so she leaves and returns a few minutes later with her supervisor.
Here is the conversation with the supervisor.

“Which doctor,” she asked, “told you to stand there?”

“I don’t know, he didn’t tell me his name.”

“How long ago,” asked the supervisor, “did the doctor tell you to
stand there?”

“It must be about five hours ago,” I replied.

“Do you expect us to believe that a doctor whose name you don’t
even know told you to wait there five hours ago and that he hasn’t come
back yet?”

I remained silent. The question was extremely sharp, and I winced
under its impact.

“Are you sure that the doctor is not just a figment of your
imagination?”

124

Part I

I must admit that my confidence was shaken. I began to doubt,
myself, that this had happened. Plainly doctors don’t break their
promises. If one of them tells you that he’ll return in five minutes, he
won’t keep you waiting for five hours. Such things just don’t happen.

“I must be crazy,” I said weakly, “Maybe it didn’t happen.”

The supervisor nodded. “That’s better,” she said grimly.

My head reeled and my legs tottered. I was beginning to feel like the
prisoner who is kept in a padded cell for years without ever being
informed of the charges against him in Kafka’s novel The Trial.
Everything was turning topsy-turvy. Guilt was innocence and innocence
was guilt. Nothing admitted of any rational explanation. If that doctor
who had told me to wait for him for five minutes were merely a figment
of my imagination, then I was losing all contact with reality. But if that
doctor really had existed, then why did all these people maintain that he
didn’t? I knew that he existed. I had seen him and spoken to him. Then
all these people must be engaged in a monstrous conspiracy to drive me
mad. No other theory seemed to fit these facts.

The supervisor interrupted my thoughts. “Come with me,” she said.

I obeyed meekly and followed her towards the center of the hallway.

“You're very confused, young man,” said the supervisor. “Because
you’re new here I won’t report how you’ve just behaved. But I warn
you, don’t do it again. I want no further trouble from you.”

Again, then, we see the use of this procedure - that one can choose among
facts according to the presence or absence of an explanation. It’s absolutely
routinely used. I’m not proposing that it’s obscene, I’m only proposing that
that’s the way it’s done. At least in this society, facts and explanations have
more than a one-way relation to each other. That is, it isn’t the case that if
something has occurred, that sets the problem ‘construct an explanation,’ but
the notion that persons hold of possible facts is that those facts are possible for
which there is an explanation. One can’t merely say “Well, I saw it. You
explain it.” Something proposed to have occurred can be treated as not so, by
virtue of the fact that there’s not an explanation for it. That’s important in this
society, given the fact that miracles are no longer usable. And miracles are that
class, in part. They’re events for which there is no account, which now
systematically would be given an account, i.e., an account not of this world.
But most persons who consider themselves to be modern individuals don’t
read, for example, the psychical researcher, Rhine. They just take it that
whatever he says, he’s a fraud in one way or another, and you don’t have to
bother coming to terms with what he reports. Given that what he proposes
as an explanation couldn’t be an explanation, then it hasn’t occurred. And
further, there is an explanation available which provides that there is no such
phenomenon: Statistical chance. Such an explanation proposes that out of a
population there will be some persons who produce these responses, and he’s
happened to have found them. But that does not provide any basis for psychic
research; random distributions would do the same job.

Lecture 14

125

That’s a classical problem. It can always be raised if you have a limited
amount of research, where, then, if you propose there’s a phenomenon, it is
said that you have nothing but the working of a random distribution, which
you happen to have caught at some point. And it’s raised more than
occasionally about psychic research, and I suppose more than occasionally it
turns out to be relevant. Or, for example, people say “How does it happen
that there have been two earthquakes on this day?’’ and then they go about
constructing some explanation: The gods are angry. Or, “Isn’t it odd that the
weather pattern has changed this year?” and then they construct an
explanation: Fallout. And now what’s proposed is that if you use such a unit
as ‘years’ to measure weather, where weather as a phenomenon has to be
measured over eons, then you’re liable to find something which looks like
order. But given the proper unit, that proposed order is not present. Now,
there are all sorts of units available, and for any proposed phenomenon, that
unit which would make that proposed ordered set of facts simply a
coagulation of random events can be proposed to be the relevant unit. Either
side may be right or wrong. Nonetheless it’s done.

Appendix A
A Note on the Editing

In lecture 2, p. 17, Sacks introduces a consideration:

A recurrent thing that I’ve seen throughout this stuff is persons talking
about not feeling wanted anymore. The question is, how is that kind of
feeling provided for in this society? And what would be interesting
about it would be if we could see some way in which, quote, the
structure of society, provided for the focusing of kinds of troubles.
That’s what I think we can see with this . . .

Throughout this volume many of Sacks’ pronominal uses have been changed.
Here, the operation is more or less innocuous. What is rendered as ‘. . . if we
could see . . .” and “That’s what I think we can see . . .’’ actually goes,
“. . . if we could see . . .” and “That’s what I think you can see . . . i.e., the

second ‘we’ is actually ‘you.’ Following is the unedited version. (Emphasis on
pronouns is always added: it is not in the transcripts.)

Now, a recurrent thing that you have probably seen, and I’ve seen
throughout this stuff, is persons talking about not feeling wanted
anymore. The question is, how is that kind of feeling provided for in
this society. And what would be interesting about it is, if we could see
some way in which, quote, the structure of society provided for the
focussing of kinds of troubles. That’s what I think you can see with this.

(Fall 1964, tape 5, side 1, p. 10, unedited)

And this ‘we’/‘you’ alternation is preserved in the abbreviated earlier
transcript (transcriber unknown).

The recurrent thing is people talking about not feeling wanted. How is
that kind of feeling is {sic] provided for in this society? If we can see
some way in which the structure of society provided for the focusing of
certain kinds of troubles. That’s what you can see above.

(Fall 1964, tape 5, side 1, pp. 3-4, transcriber unknown)

Another ‘you’/‘we’ alternation occurs just a bit further in the discussion. Here
is the unedited version.

Now you see that what she’s picking up here is, quote, the absence of
ceremonials. Now, we see that ceremonials have this double use.

(Fall 1964, tape 5, side 1, pp. 10-11, unedited)

126

127

Appendix A

(Neither the edited version nor the earlier transcript show the repeated,
pronoun-alternated reference to ‘seeing.’ The edited version has it:

What she’s picking up here is the absence of ceremonials. And
ceremonials have this double use.

(Fall 1964-Spring 1965, lecture 2, p. 18)

And the earlier transcript goes:

She picks up the absence of ceremonials. Ceremonials have a double use.

(Fall 1964, tape 5, side 1, p. 4, transcriber unknown)

So, in close order there are two of these pronominal alternations: “. . . if we
could see . . . ‘/’I think you can see” and ‘‘Now you see . . .’’/“Now, we
see ...” And such alternation is recurrent across the lectures. Following are a
few instances, taken from the Fall 1964 and Spring 1966 unedited lecture
transcripts.

Now, what you can see is this . . . And what we find is . . .

(Fall 1964, tape 1(R), p. 6, unedited)

This cross-paragraph alternation survives in the edited version (Fall 1964-
Spring 1965, lecture 1, pp. 6-7).

If you look at [X and Y} what we seem to find in them is . . .

(Fall 1964, tape 4, side 2(?), p. 1, unedited)

This alternation is edited out and rendered as: “What we find in [X and Y]
is . . .” (Fall 1964-Spring 1965, lecture 7, p. 54).

I could say [X is the case] but / wouldn’t be able to show why that’s so,
or how it’s so. If that’s the case, I don’t establish my point. I could say
[Y] and we could [find A but not B}. That again wouldn’t do what we
propose to do. We want to do both: [A and B], You want a method that
generates this.

(Spring 1966, lecture 4, pp. 1-2, unedited)

In the edited version, a ‘we’ referring to ‘I’ is changed to ‘I’ (“That, again,
wouldn’t do what I propose to do”), and a ‘you’ following a prior ‘we’ is
changed to ‘we’ (“We want to do both: [A and B ] ... we want a method that
generates this line. (Spring 1966, lecture 4, p. 301).

In short, what you want to do is [X]. And we can take it that [Y],

(Spring 1966, lecture 21, p. 6, unedited)

This alternation survives in the edited version (Spring 19 66, lecture 21,
p. 420).

128

Part I

If you can claim that it could be another thing, first one wants to show
how it’s another possibility.

(Spring 1966, lecture 22, p. 2, unedited)

This alternation is revised to, “If one is claiming that it could be another thing,
first one wants to show how it’s another possibility’’ (Spring 1966, lecture 21,
p. 422).

Many of the pronominal alternations are edited out or revised. One major
basis of revision is an attempt to deal with the ‘direct address’ problem.

Sacks often addresses remarks directly to the class, as occurs in the unedited
transcript of the paragraph in question here:

Now, a recurrent thing that you have probably seen, and I’ve seen
throughout this stuff . . .

He also makes such meta remarks as, “I would like very if you could collect
instances of the uses of {X}” (Fall 1964, tape 4, side 2, p. 11, unedited).
‘. . . if it isn’t clear, stick up your hand and tell me’’ (Fall 1964, tape 1, p. 2,
unedited), not to mention things like “Oh by the way. Somebody tell me
when I get - I don’t have a watch. So just keep me vaguely informed about
( )” (Fall 1964, tape 1, p. 9, unedited). These have been edited out.

But also, lecture-relevant talk which was addressed to the class was often
deleted or changed because it could be troublesome to readers, either
estranging them, making them onlookers or overhearers, or over-intimatizing
the talk, seeming to address this particular reader directly. So, e.g., such a
remark as, “You ought by this time to be quite aware of the fact that
(Spring 1966, lecture 6, p. 1, unedited) was changed to “By this time it
ought to be quite obvious that . . . ’’(Spring 1966, lecture 6, p. 312). In the
actual situation, of course, members of the class were neither overhearing, nor
being individually addressed by, such utterances.

There are, then, variously problematic uses of ‘you.’ And the fact that
Sacks not infrequently uses ‘you,’ ‘one,’ and ‘we’ in alternation became a
resource - and license - for changing them. Most often ‘you’ was changed to
‘we,’ a word Sacks makes extensive use of, sometimes in rather idiosyncratic
ways. So, for example, he will use ‘we’ where he himself is the referent, e.g.:

That again, wouldn’t do what we propose to do.

(Spring 1966, lecture 4, p. 2, unedited)

Let’s start out with something we mentioned before . . .

(Fall 1964, tape 14, side 2, p. 2, unedited)

Remember we said about the opening lines of conversations that they
seem to come in pairs.

(Fall 1964, tape 5, side 1, p. 3, unedited)

129

Appendix A

Now we originally introduced the notion of an omni-relevant device . . .

(Fall 1966, lecture 4, p. 6, unedited)

in each case it being he himself who has “propose{d},’’ “mentioned,” “said,”
and “introduced [a] notion.”

In one case, Sacks recasts his own utterance:

We haven’t said that they have a right to be told. I have not said that
they have a right to be told some trouble of A’s.

(Fall 1964, tape 10, side 2 (M3 side 2), p. 11, unedited)

And in the edited versions, either ‘we’ is changed to ‘I:’

That, again, wouldn’t do what I propose to do.

(Spring 1966, lecture 4, p. 301)

I’ll start out with something I’d already worked up . . .

(Fall 1964-Spring 1965, lecture 8, p. 57)

I said about the opening lines of conversations . . .

(Fall 1964-Spring 1965, lecture 2, p. 13)

or the talk is revised to accommodate ‘we:’

We have now, the notion of an omni-relevant device . . .

(Spring 1966, lecture 6, p. 316)

Sacks also uses ‘we’ to refer to pretty much anyone but himself, e.g.:

Hopefully the thing I want us to see in this first is . . .

(Fall 1964, tape 5, side 2, p. 1, unedited)

I raise this because while we all can see that that’s quite so, there’s a
related and in a way much more interesting thing that I doubt we’ve
noticed . . .

(Spring 1970, lecture 1, p. 7, unedited)

We tend somewhat to be perhaps overly taken with the constraints that
Weber sets for an objective social science . . .

(Spring 1966, lecture 33, p. 9, unedited)

where it is members of the class, and not Sacks, who will “hopefully” come
to “see” something, and whom he can “doubt” have “noticed” something.
And it may be an even larger population, but in any case one that doesn’t
include Sacks, whose members “tend to be perhaps overly taken with”
Weber’s constraints.

130

Part I

While these self-inclusive, ‘we’-references may be incorrect in a strict sense,
they may be doing the sort of work Sacks discusses in one of the lectures; that
of saying ‘everyone.’

If someone who needn’t include themself in some class . . . wants to say
in a stronger way than ‘you’ (which can be misheard as only the
recipient) that ‘everyone is that way,’ they can say ‘we.’

He offers a case in which a psychiatrist remarks, “We can be very blind to the
things around us,” and goes on to say:

That is, by virtue of the sequence in which ‘we’ is heard, the inclusion
of the speaker is partially gratuitous; is heard as doing something he
‘needn’t have done.’

(Spring 1966, lecture 11, pp. 351-2)

As it happens, in the edited versions those instances of “gratuitous”
self-inclusion in the environment of a demand, “. . . I want us to see . . .,” or
criticism, “. . . there’s ... a much more interesting thing that I doubt we’ve
noticed” and “We tend somewhat to be perhaps overly taken with [X].
were preserved. (See Fall 1964-Spring 1965, lecture 3, p. 21, Spring 1970,
lecture 1, p. 00, and Spring 19 66, lecture 33, p. 487.)

In contrast to those pronominal changes mentioned earlier, both the
changes from ‘we’ to ‘I’ when Sacks himself is the referent, and the
preservation of the self-inclusive ‘we’ in the environment of touchy assertions,
were not matters of policy; they were decided upon one at a time. But they
can be seen to add up to something like policies, or perhaps to reflect
underlying policies.

The paragraph at the start of this consideration, plus a bit of subsequent talk,
happen to instance another phenomenon that was often revised. The unedited
transcript shows a format, “quote, [X},” which occurs once,

... if we could see some way in which, quote, the structure of society
provided for the focussing of kinds of troubles . . .

and is then used again some 16 lines later.

. . . what she’s picking up here is, quote, the absence of ceremonials.

(These two fragments are from Fall 1964, tape 5, side 1, p. 10, 11. 4-6 and
p. 10, 1. 23-p. 11, 1. 1, respectively. See p. 126 of this Appendix for fuller
fragments of each.)

This phenomenon of a repeated format occurs throughout the lectures, and
often - as here - one of them, usually the second, is deleted in the edited

Appendix A 131

version, for both aesthetic and technical reasons. Only the latter will be
considered here.

One sort of technical problem posed by the repeated format is that a term
which may appear to be analytically based, can be merely format generated.
And a glimpse of that problem may be gotten from the recurrence of the term
‘quote’ in “. . . quote, the structure of society ...” and ”... quote, the
absence of ceremonials.” Roughly, the quote-marking of the familiar phrase
‘the structure of society’ seems to be doing its standard ‘as they say’ work,
where a new way of working with an old concept is being offered. It is not
clear what sort of work the quote-marking of the newly minted technical
descriptor ‘the absence of ceremonials’ is doing. One could not in this case
substitute ‘as they say’ for ‘quote.’ It may simply be a format-recurrence. In
any event it was treated as such, and deleted from the edited version which
simply goes: “What she’s picking up here is the absence of ceremonials” (Fall
1964-Spring 1965, lecture 2, p. 18).

Part II
Fall 1965

Lectures on Conversation, Volume I, II Harvey Sacks
© 1995 The Estate of Harvey Sacks. ISBN: 978-1-557-86705-6

Lectures 1 and 2

“The baby cried. The mommy picked

it up.”

Lectures 1 and 2 were not recorded , but Sacks had made preparatory notes for the
“first lecture(s)" which seem to have turned out to cover lecture 1 (see Appendix
A). The second lecture is covered in notes Sacks made after its delivery (see
Appendix B).

Lectures on Conversation, Volume I, II Harvey Sacks 135

© 1995 The Estate of Harvey Sacks. ISBN: 978-1-557-86705-6

Handout Group Therapy Session

Segment

The door opens in the middle of a group therapy session, and in comes a new
entrant. Until now, at this session, the members have been three teen age
males, and a male therapist of around 35 years of age. The new entrant is a
male teenager.

Time

0:00 Therapist'.
Bob:

Th

Bob

Joe

Th

Henry

Bob

Th

Joe

Henry

Mel

Joe

r Mel

- Henry
-Mel
L Henry
Joe

1:00

Bob this is uh Mel

hi

Joe

hi

hi

Henry

hi

hi

Bob Reed

(cough) We were in an automobile discussion,
discussing the psychological motives for
drag racing on the streets.

I still say though that if you take, if you take uh a big

fancy car out on the road and you’re hot roddin’ around

you’re you’re bound to get, you’re bound to get caught

and you’re bound to get shafted. We- look

Now did you do it right. That’s the challenge That’s the

challenge You wanna try

That’s the problem with society. Hahhh

And do it right so you do not get caught.

That’s the

In that Bonneville of mine? I could take it out with me
and if I got a tie and a sweater on and I look clean? 99%
of the time a guy could pull up to me in the same car the
same color, the same year, the whole bit, roll up his pipes
and he’s in a dirty grubby tee shirt and the guy will pick
the guy up in the dirty grubby tee shirt before I before
he’ll pick me up.

Just - just for -

This is as close a copy as possible of Sacks’ transcript.

136

Mel

Joe

Henry

Joe

Henry

Mel

Henry

Joe

Mel

Henry

Joe

Mel

1:30 Henry
Mel

Joe

Henry

Joe
Bob
Mel
Bob
Henry
Mel
Bob
Mel
Bob
Mel
Bob
Joe
Mel
Bob
Mel
Henry
Henry
Mel
2:00 Bob
Mel
Bob
Mel
Bob
Mel
Bob

Handout: Group Therapy Session Segment 137

Not many people get picked up in a Pontiac station
wagon.

Now I agree it looks like a daddy’s - It looks like a damn
mommy’s car.

Joe, face it. You’re a poor little rich kid.

Yes Mommy.

Face the music

Ok Now you’ve got that out of your system. Now you’re
a poor little rich kid we’ve told you that.

And we also decided you’re a chicken shit.

I decided that years ago. Hell with you

Now let’s see what else can we decide about you?

Hey don’t tear him down
I’ve been torn down for-
ok

We got company

oh ok. Tell us all about yourself so we can find something
bad about you.

Yeah. Hurry up.

Well first of all you must be crazy or you wouldn’t be

here.

heh heh

Yeh I guess

Secondly, you must be an under-achiever
yea

You hate your mom and dad, huh
Third of all —
oh sometimes

Fourth, you like to drive cars fast,
yea

Fifth you like uh you like wild times
Yeah

He smokes like me see.

Sixth, you like booze
yeah

Seventh you like to smoke

And seven you’ve been arrested for rape and other things.

Ha

Eight you-
No. not that.

Eight you Eight you give lip back to everybody.

Yea

Nine you uh cut classes,
yea

Ten, you’ve been kicked out of school once,
yea

138

Part II

Mel

Joe

Henry

Mel

Henry

Bob

Henry

Bob

Mel

Henry

Bob

2:30 Mel
Bob
Henry
Bob
Mel
Joe
Bob
Henry
Bob
Henry
Bob
Henry
Bob
Henry
Mel
Henry
Joe
Henry
Joe
Henry
Joe

3:00 Henry

Joe

Henry

Joe

Henry

Henry

Joe

Joe

Mel

Bob

Joe

eleven uh
hah hahh

We’re doing better than he is. haha. Proceed, I’m with
you. ha ha.

Your turn.

Are you just agreeing because you feel you want to or

what?

huh?

You just agreeing?

What the hells that?

Agreeing

Agreeing

Agreen?

Agreeing. With us. Just going along with it.

No.

Saying Yes. Yes. hahhh

Its true, heh, everything you said is true. so.

You wouldn’t be in this group if you didn’t.

What school did you go to?

I went to Palisades
Went?

Yeah. And then University.

Oh I went to Uni. I graduated there.

You did?

Bitchin school isn’t it
Yeah I guess

Yeah I never did ( ) it

What did you do with the matches?

You don’t like it do you?

I got them
Here you go chum
Everybody’s faster than me today
You got reflexes like a slow turtle
God damn you ha

Remind me later. I’ll show you a litte game about your
reflexes.

Got any more cigars.

No it’s last one
Last one
Pity heh

Usually there’s a broad in here. Pretty -

Her name’s Barbara. She pretty good. She’s real nice.

(cough)

A lot of her to love but
Sounds like an old man

Huh. This is the old, this is the couch society of what is it?

139

Mel

Bob

3:30 Mel

Joe

Mel

Henry

Joe

Henry

Mel

Henry

Mel

Joe

Bob

Joe

Henry

Mel

Henry

p

Henry

Bob

4:00 Henry
Joe
Mel
Henry
Joe
Henry
Bob
Henry
Bob
Mel
Joe
Mel

p

Henry

Bob

Henry

Bob

Henry

Bob

Mel

Henry

Bob

Joe

Handout: Group Therapy Session Segment

Wanna join our army?
hmmm?

Wanna join our army? We’re gonna go defeat the king of
Greece.

We’re gonna

The academic counselling center fightin’ army’s gonna go

Alright Mel, he’s been here ten minutes

Mel. Your match is still going Mel

Mel He’s been in here ten minutes let’s judge his

character.

ok. ready Henry?

ok.

You’re first

You’re gonna get killed
ha

Keep your mouth shut and don’t say a damn word, ha ha
He’s not at all like Joe.

No

He’s more like Mel and I

( )

What you refer to as - hippie, ha. He’s been in it up to his

neck

yeah

A couple of times
Ha ha. Taste good?

No. I bit off the end of it. I was chewing the end of it.
umm.

Bitchin

Havin a big hassle with your folks
yup

Right. Daddy wants to keep you down Mommy
hmmm?

Mommy you can tell what to do
(cough) God damn cancer

See. I told you, we know everything about this guy already
Sure

Do you have any brothers or sisters?

No

Daddy’s sort of sheltered you

No not really

No?

Not really

You sorta shelter them?

Do you have a car?

No

(gasp)

Part II

140

4:30 Mel
Henry
Bob
Henry
Bob
Mel
Henry
Bob
Henry
Bob
Joe
Henry
Bob
Joe
Mel
Henry
Bob

pause

5:00

Mel

Therapist
| Henry

r Henry
Joe

Henry

Mel

Bob

Joe

Henry

5:30 Bob
Mel
Joe

Henry

Henry

Mel

Henry

vMel
-Henry
L Mel

oh oh. One point against you.

You had a car?

No

You want a car
oh hell yes, ha.

You’ve stolen a car?

Daddy doesn’t want you to have one.

No he wants me to have one, my old lady doesn’t.

oh!! Main conflict with mother.

yes

hehh

Mother’s too dominating,
yes

But you’ve been had already fella, ha ha
ok

Bring your mother in, we’ll work on her. ha ha
ok

What are you thinking, old man?

hahh I’m thinking Bob has certainly had it. hahh

Whip out your pad

Five minutes and you guys

No No write down what we say take the ( )

You gotta pardon these guys they ’ re all they ’ re all the

social outcasts of the world.

This one’s chicken shit. He’s a bastard, and I’m a Mau
Mau.

ok. good. Now we’re all settled, what are you?
oh I’m
You a hood?

This is an abnormal session. See we’re not together without

the broad.

yeah

See we gotta have the broad here. Cause she unites us.

She unites us - she keeps us on the road. Ha ha. Table’s
always ready and ha —

She sorta keeps us on the road
oh well

oh well you know that the way things go. That’s the way
things lie down.

He sorta keeps his mouth shut and writes down things
whenever you say something important
And then if you ask him
He’s a good guy though

to do something then you have to pay him though

141

Bob

Henry

Joe

6:00 Henry
Bob
Henry
Mel

Henry

Handout : Group Therapy Session Segment
yeah

Eventually you’ll become sane,
yeah but you —

Or your mother will which is not what you think.

I hope so

You’re not crazy, it’s your mother, hah?

But uh it takes about, I’ve been here four months and it’s
getting better now. I used to think that I was a bird and
would fly.

Yea He used to walk out on us, he thought he was above

Mel

Henry

Mel

Joe

Henry

Bob

Joe

6:30 Henry

Mel

Mel

Henry

Joe

Mel

Henry

Bob
Joe
Mel
Henry
Mel
7:00 Joe

Henry

Mel

Henry

Joe

Mel

Joe

Henry

Yea. But now I’m now I’m below you.

Yeah. I corrected I corrected that quality. I gave him an
inferiority complex, ha.

And I got him to shave,
hehh.

Yeah. I’m not grubby or nothin
No. hehh

hah. Hey this is the academic counselling center. It’s called
the family, family circle.

It’s not really an academic counselling center; it’s sort of a
drive in nut house, ha ha.

Then you rent a counsellor; $ 5 an hour and 50<t a question,
ha ha.

Don’t consider this one- but on the whole you’ll enjoy
’em. Ha. In fact it’s a good way to spend Saturday morn-
ing,
hahh.

After Friday night.

The morning after the night before. Ha ha. After a couple

cups of coffee you’re alright.

yeah

A couple dozen cups of coffee.

Chew a cigar now and then and
I don’t smoke (he’s been smoking a cigar)

He just started last night

You gotta believe it. You gotta believe it we’re the only
two in here that smoke other than him.

Doesn’t Barb smoke?

No. she doesn’t smoke. She said she used to.

She’s a good girl, ha ha

She smokes cigars. That’s about all. ha ha.

They smoke. I chew them.

He eats it.

Well chewing them is half the fun - oral gratification, heh
heh

142

Part II

Mel

Henry

Joe

7:30 Henry
several
Henry

Mel

Henry

several

Henry

several

Mel

Henry

Mel

Henry

several

Mel

8:00 Henry

Joe

Mel

Henry

Mel

Henry

Joe

Bob

Mel

Bob

Henry

8:30 Joe
Mel
Henry
Bob
Mel

Let’s see, what else can we tell him?

They say that’s why bubble gum and all that shit’s good uh

so popular

Why?

There’s a trend to put things in your mouth - like thumbs
(laughter)

No really. Why do you think babies suck their thumb and
adults chew gum and smoke? Everybody wants to keep
something in their mouth.

Well, you know -

Now I’m not talkin about eating or nothing. And I’m not
talking about what you’re thinking of either. Hahh.
(laughter)

No really I read this in some book

(laughter)

oh come on

It’s really the truth

one more thing —

Look at him. Look at him sucking at that pipe all goddam
morning.

(laughter)

He sucks on his dentures too. Now listen, one more thing.
There’s people watching us now. They got a tape recorder
and

Thai is the microphone (pointing). Occasionally
there’s somebody in there judging us. They’re baby
headshrinkers. They’re learning how to be.

Don’t worry about ’em. They destroy the tapes afterward.
He’s not quite one. He won’t be one until two months.
He’s writing a -

paper. But today we’re not being uh viewed. We’re just
being taped.

No. we’re being viewed. There’s somebody in there. Cause
they lit a cigarette.

But he’s blind.

Hey. Don’t worry about it.

How how could you see him?

Well he can see us but you can’t see
Can you see through this here?

When you turn the lights on. See it’s darker in there. I
think -

Wave at him. He’s a nice guy.

Anything else Henry?

And if it gets private we can always shut the curtain,
heh great,
to do something.

Mel

Joe

Henry

Mel

Henry

Mel

Henry

Mel

Joe

Mel

Henry

Joe

9:00 Henry

Handout: Group Therapy Session Segment 143

Anything else Henry? D’you think we oughta let him in
on? or tell him?

We’re all nuts?

Naw. Let him slide for a few weeks,
ok

Get the hang of things, heh.

Then we’ll really go after him. Then when we - when we

get -

We ll let you rest with your problem now

We’ve told you all your problems now you gotta rest with

them.

Yeah

Then we’ll tell you how to solve them.

You gotta get warmed up
(cough)

And vice versa. 1

1 On the original transcript, the materials fitted precisely onto six pages. The tape
goes on for another hour and 40 minutes, or so. Whether the fragment would have
continued on until the bottom of a next page, or have cut off at this neat
time-and-topical place, is an open question.

Lecture 3

A collaboratively built sentence; The

use of ‘We’

I’m going to begin discussing this eight-minute segment from a group
therapy session for teenage kids that I gave out last time. When I used this
last year as a final exam, I offered as a suggestion for considering it, that I took
it to be an initiation ceremony, and as such, among the things in it that are
of interest is the fact that it isn’t done via a script. I think it may be the case
that one of the participants in the group knew beforehand that somebody new
was going to come that day, but that’s about it. There isn’t any basis at all for
supposing that this thing was planned by them. And yet (if ‘yet’ is at all a
relevant term; it’s relevant vis-a-vis the literature on initiation ceremonies
anyway) there are certainly features in it which are proposed to be present in
initiation ceremonies. But our task is not going to be so much to see the extent
to which that is so, but to try to come to terms with this thing, as best we can.

I’m going to begin consideration of it with what I take to be Segment 2,
which runs from the cough which is right after the introduction of the various
personnel, until that point which is very close to the 1:30 marker, where Mel
says “Oh, ok.’’ So it runs from “cough” to “oh, ok,” inclusive. That’s going
to be the subject-matter for these first considerations. Perhaps we’ll eventually
move backwards and forwards.

There are, in the conversation itself, a lot of events that are to the altogether
naive eye, quite remarkable. That is, without any analysis and simply by
inspection you can find some things that you might take to be worth thinking
about, without any special consideration of what we’ve done here at all.
Probably the most striking thing right off, would be that part of the third
section where there’s a series of eleven questions run off in about 45 seconds,
by the given members of the group to the new entrant, in which they ask him
things about himself, giving in each of the questions what they take to be the
answer. And where the questions have some rather extraordinary detail in
them, for example, question ten, “Ten, you’ve been kicked out of school
once.” The issues of how these guys have that information, and what they do
with it, are quite obvious ones. And perhaps we’ll eventually get to consider
that.

Segment 2 begins with what is, in its way, also a rather remarkable
occurrence; and that is, the three boys collaborate to produce a single sentence:

Joe : (cough) We were in an automobile discussion,

Henry : discussing the psychological motives for

144

Lecture 3

145

Mel : drag racing on the streets.

It’s not altogether unheard of for two persons to collaborate to produce a
single sentence. The normal way that is done, however, is that, say, one person
produces an almost complete sentence and finds himself searching for a last
word or a last phrase which he can’t find, and the other offers it. Another place
where one gets something like this is in the following sort of thing: A person
makes a request - ‘Would you pass the salad.” Somebody else, maybe the
person who got the request, then appends to the request some word or phrase
that they propose properly completes it. . .((voice from the gallery: “Please”))

. . .“Please;” which they propose that, lacking that, the request is not
complete.

Now those are two sorts of things worth considering in their own right. But
they’re far off from this thing we’re looking at. It’s a kind of procedure of such
power that if it were used as a literary technique it would just be rather vulgar,

I suppose. I know a couple of literary uses, though I can’t really pin them
down for sure, and if anyone knows of any you could tell me sometime, or
collect them yourself. 1 In a musical version of Edgar Lee Master’s Spoon River
Anthology he uses this as a way that the pair of lovers talk. 2 That is, the pair
of lovers talk in the following sort of way: Each one produces part of a
sentence, which the other then may complete. So you write out a sentence,
break it somewhere, and have the other part of the pair continue it.

Now Masters was writing a while ago, and I take it that we would take it
that that’s an altogether obvious device to show, through this playing with the
syntactic features of an utterance, that these people are close to each other.
They’re a unit. Because a sentence is obviously a prototypical instance of that
thing which is done by some unit. Normally, some single person. That then
permits it - for those who have the wit to do it - to be a way that some
non-apparent unit may be demonstrated to exist.

We get, then, a kind of extraordinary tie between syntactic possibilities and
phenomena like social organization. That is, an extremely strong way that
these kids go about demonstrating that, for one, there is a group here, is their
getting together to put this sentence together, collaboratively.

It’s hard to figure how they could do that right off, in anything like as
sharp a way as they picked. As it happens, as rare as I take it that this kind
of a thing is - and until I saw it I don’t think I had seen anything like it before
- in the very same segment there’s another. Between 1:00 and 1:30 we get
the following sequence:

Mel : Ok Now you’ve got that out of your system. Now you’re a poor

little rich kid we’ve told you that.

Henry : And we also decided you’re a chicken shit.

1 Written in at the margin: “Donald Duck. Huie, Louie & Dewey.” A student
contributed this.

2 Written in at the margin: “cf. Freud.”

146

Part II

In the case of the participle “discussing,” which constructs what is being
said as a dependent clause and makes, then, the statement before it a clause
also, the phenomenon is rather sharper than this one, where “and” is a rather
normal way to begin a sentence. But it seems to do exactly the same kind of
job, for, now a different group. That is, if the first group (Joe, Henry, Mel)
is formed through their production of “We were in an automobile discus-
sion . . . ,” etc., and is demonstrated to be present through that production,
as an organized entity in this terribly powerful sense of being organized in
such a sufficient way that a task properly done by a single person is done by
all of them together - and not as a chorus - then, there are clear ways that
Mel’s “Now your’re a poor little rich kid we’ve told you that,” followed by
Henry’s “And we also decided you’re a chicken shit” is very similar. I’ll
consider this one when we get down that far into the conversation. And if you
examine the conversation with the slightest attention, certainly one would
take it that at that point in it a subgroup is being formed, consisting of the
two people (Mel, Henry) who collaborate on that sentence.

Such mobilizations of syntax are normally assigned to purely literary
phenomena - poetry, the novel, the play - as their expectable environments.
That they might be explored by laymen doing quite unformalized tasks is
something that is, I take it, quite worthy of being investigated - for other
things than just this particular way that these syntactic features are being
handled.

Now let’s see to some extent how that thing is done. Joe’s utterance, “We
were in an automobile discussion,” is something which by itself could be a
perfectly decent sentence. And it is, then, in the first instance, what Henry
does that makes it the independent clause of a sentence which is now to have
a dependent clause. That is, Henry makes his own statement not possibly a
sentence by using the participle “discussing,” which is just not a way of
beginning a sentence . 3

About the third part (Mel’s “drag racing on the streets”) there is no
question that it collaborates with the second in making of the first, ‘the
independent clause.’ Neither the second or third alone are sentences, and the
two together do not make a sentence. Only with the first is it all a sentence.
So that particular choice of participle is to be accounted for by reference to
some task of social organization, solved by reference to syntactical features.
And not by any, quote, purely linguistic considerations. Or even stylistic
considerations. The participle, then, becomes an object in the technology of
social structures, I suppose. And we get some work that it can do which one
isn’t going to find much in whatever radical grammar you’re going to look
at . 4 That would be trivially true for almost every grammar, since they would
never make a distinction in the first instance, between speakers. And yet it’s
obvious that, that this statement is made by three persons makes it quite a
different statement than if it was made by one. And if you took it as all made

3 Written into the margin: “Needs to be modified, for it can begin a sentence.”

4 Written into the margin: “And we don’t find this in sociology texts either.”

Lecture 3

147

by one, the various ways that they might have of doing the kinds of things
they can do when they make it together, would be enormously weakened.
Suppose, for example, Joe had said the whole thing. And the two of them
either said “Yeah,” or nothing. The character of what they would be
affirming would be possibly obscure.

So it’s not only that they present themselves and demonstrate that they are
organized, but by collaborating on the sentence they have this extremely
powerful way of showing that they do agree as to what it is that the topic was:
They all have some part in saying what it was. And I take it that that’s a
different kind of technique of affirmation than is, A says the topic was X, and
B says yes it was. And that would be especially relevant where the question
is, what does that topic mean; where, for example, a person who hears it may
take it that he is not sure, and perhaps neither of them really agree on it.

In the way that these fellows fit their talk together, which would involve
a hearer in seeing that they ‘know what’s on each other’s minds,’ he could
presumably as well take it that what it is they’re talking about, they also
know. And that the thing works this way would of course be further
evidenced if we had more studied uses of it; that is to say, literary uses, for
example, where it is obviously just that point that’s being made. That is to
say, what better way is there for lovers to show that they are one . 5

In that regard it’s perhaps worth opening up one’s notion of what it is that
organization might be directed to doing. There is, I take it, a considerable
tendency to think of the task of organization as being the solution, by a set of
persons, of some job which no single one could, or could efficiently, do.
Although that is certainly so to some considerable extent, the fact that there
is a job that any person could clearly do by themself, provides a resource for
members for permitting them to show each other that whatever it is they’re
doing together, they’re just doing together to do together. That is to say, if one
wants to find a way of showing somebody that what you want is to be with
them, then the best way to do it is to find some way of dividing a task which
is not easily dividable, and which clearly can be done by either one alone. And
that, I take it, obviously is done rather frequently, if not with the putting
together of sentences, then with all sorts of other things. Where neither party
can then readily take it that the other is simply being of help, but is involved
in seeing that what they’re doing is ‘doing whatever you’re doing with you.’

Now there’s certainly ambiguity in ‘divisible’ as a term. And what I’m
pointing to is not something which is not readily divisible and is also
unmanageable by one person, but something which is not readily divisible but
which is manageable by any single person. The point can be made by
paraphrasing a fantastic aphorism of Kafka’s. He loved to play with the
problems of social organization, and give them queer formulations. And at
one time he’s considering Neptune’s problems. Neptune is the bureaucrat
who runs the oceans, according to Kafka. And he has the following kind of
problem: How in the world is he supposed to divide up that kind of work?

5 Written into the margin: “cf. Freud”

148

Part II

You can’t, after all, assign somebody an ocean. It’s not that sort of
non-divisibility that we’re talking about.

Now there are, aside from what I’ve just mentioned, a variety of other
things involved in this segment that we’re going to have to consider in one
way or another.

An extremely sharp formulation is made of the term ‘we,’ where it remains
the subject of the sentence without each person affiliating himself to it. And
we might then say about ‘we,’ that what it refers to is ‘those three guys.’ That,
however, would be a possibly troublesome kind of way of characterizing it.
For what we might have to decide, then, is what kind of form of ‘those three
guys’ does it refer to? That is, is it a list in which they happen to be the only
people on the list - that is, directly pointed to by ‘we? They talk in a sequence
which might permit one to say they are a list of persons, and previously they
are introduced as a list; a list of nameables. Now if it is a list, then there are
of course a variety of kinds of operations which might be permissable on that.
For example, order might not matter at all. If we wanted to write them down
we could say ‘we:’ (Mel, Henry, Joe) or (Joe, Mel, Henry), it doesn’t matter.
As long as you have them there, that’s it. At least for some purposes it would
seem intuitively obvious that order is not irrelevant to the use of those names,
and if we go back to the first segment, we can see such a one.

When the new entrant comes in, the others are sitting around the room, in
chairs. There is the therapist, then Mel, then Joe, and Henry. Bob comes in,
and he’s introduced. There’s a pause in the introduction, right at the
beginning: “This is uh Mel.” Now we can take it that at that point matters
are open as to how they’re going to be introduced, i.e., in what sequence. And
we can take it that probably only two alternatives are present: Go around to
the therapist’s left, or around to his right. But you wouldn’t start in the
middle, i.e., with Joe. Once the therapist has named Mel, presumably
anybody who came in and knew that Mel had been introduced, could do the
job of introducing. And if it were done in any other sequence, we would take
it that something special is being done. For purposes of introduction, then, if
the persons are in some way physically arranged with respect to each other,
then we would take it that it isn’t simply an open matter as to what sequence
they ought to be introduced in. There may be some openness to it, but
probably, given the first name, there’s a proper further sequence to the end.

Now the problem about ‘we’ that we have to deal with, which is only
roughly posed by the issue of orderability, is the following. Is it the case that
‘we’ is some collection of these guys’ names, directly? Or is it some category
or a set of categories for which these fellows are incumbents? Where, then,
their names are usable to refer to them, but there is some organization,
perhaps, of those in that category. Where that organization is what’s
important. And saying that, then we have to turn to it first. Do we have to
build a category here? Or can we turn simply to the list of names as the thing
being referred to?

‘We’ clearly can refer to a category, which has as one of its crucial properties
that no intention exists of listing the incumbents, and furthermore they’re not

Lecture 3

149

listable. That is, ‘we’ can refer to an infinite population. For example,
‘Americans.’ That’s an obvious infinite population. People on a bus are an
infinite population, insofar as you don’t say ‘at time T.’ ‘We’ is stable over the
use of ‘Americans,’ and therefore it might be only incidental that these people
are listed.

Now, “What’s the difference?’’ may be a question for you. If it were the
case that ‘we,’ in this instance, is properly used as the substitute for a listing
of the personnel and could not be used unless a list could be made, and a list
were intended, then it’s clear that ‘we’ must get a tremendously more
restricted use than it in fact has.

One way we could go about approaching the issue of, is there some
category that is being referred to by ‘we’, is to consider what may be a same
use of the term in this segment - where there are quite different uses of it,
clearly. That is, there are some uses in the middle which perhaps refer to a
group whose incumbents are two persons. But the last use of ‘we’ in this
segment, in the sentence “We got company,” looks to me, anyway, like a
similar use to this first use. And now we’ve got an alternative category to the
unknown category we’re dealing with, which is ‘company.’ Whoever ‘we’ are,
‘company’ is something in alternation to that. And ‘company’ is a kind of
interesting term. First of all, it doesn’t discriminate either as to single or plural
personnel, and it is clearly a category that is in alternation to some other
category. Now it’s interesting, for one, that it’s a category in alternation to
some other category, but there are a variety of uses of ‘company’ which stand
in alternation to a variety of categories. It is, in a way, a general alternation
category. That is to say, this could be ‘family’ and that could be ‘company;’
this could be a variety of other things and that could still be ‘company.’

Lecture 4
Tying Rules

Last time I ended in the middle of a consideration of that first use of ‘we,’ and
I made some issues about the kind of term it might be. I’m not going to
continue with that now, because I’d rather hold off on considerations which
are somewhat deeper than what I want to be doing, which is just going
through some of the readily observable features of this segment. And when
we’ve gone through those, we can deal with such matters as the ‘we.’

So let’s proceed now by first considering some of the ways beyond what
we’ve so far gotten, that the utterances of this segment are tied together. In
doing that we will of course - ‘of course’ because that’s the way the data
happens to run — be re-observing things that we’ve observed earlier in the
course. So let’s get at some of the ways that the parts are tied together - where
that they are tied together is a fact, and would be part of the warrant for
saying in the first instance that there’s a conversation going on, and perhaps
even that what I picked out is an isolatable segment of it, i.e., that we have
some way of deciding that this is ‘a segment of conversation.’

When I earlier introduced some sequencing rules I said about them that
they could operate within a sentence, across sentences, and indeed across
utterances. And we have a variety of them so operating here. We can stick
initially to those that operate across utterances. They tend heavily to operate
across consecutive utterances; that is, between two utterances, one of which
follows the other.

Now right off is what looks to be an instance of this Verb Followed by
Pro-verb Rule:

Joe : I still say though that if you take, if you take uh a big fancy car out

on the road and you’re hotroddin’ around you’re you’re bound to
get, you’re bound to get caught and you’re bound to get shafted.
We- look

Mel : Now did you do it right.

In this instance the rule would deal with “are hotroddin’,” Mel’s “do” being
the Pro-verb replacing, standing for, substituting for, referring back to Joe’s
“are hotroddin’.”

There’s at least one, and surely more, Noun Followed by Pro-noun:

Mel : Not many people get picked up in a Pontiac station wagon.

Joe : Now I agree it looks like a daddy’s- It looks like a damn

mommy’s car.

150

Lecture 4

151

That is, “Pontiac station wagon” in one utterance, “it” in the next utterance.

The character of these as rules in these cases will have to be examined, and
I’ll try to say something about that in a while.

One that recurs several times, and that’s rather more complicated, we
could start out by examining in the sequence:

Henry : And we also decided you’re a chicken shit.

Joe : I decided that years ago.

The term I’m interested in there is Joe’s “that.” There are several others like
it. In Henry’s first remark of the sequence he says “That’s the problem with
society,” and “that” presumably refers back to the long if-then clause of
Joe’s, “if you take a big fancy car out on the road. . .” etc., etc. And there’s
one of the same sort, internal to a particular utterance, Mel’s “Now did you
do it right. That’s the challenge.”

I stated initially that ‘that’ is a more complicated problem. That is so
because, for one, ‘that’ can have a variety of other uses, and does, in the very
conversation we have here. For example, Joe’s “I still say though that if you
take. . .” right at the beginning of the segment, is a rather different thing.
What that involves is that we have to have some way of locating the class of
that,’ apart from the sheer occurrence of ‘that,’ which is possibly relevant to
this rule. Where, let’s say for now that the rule works something like:
Predicate ■*— ‘That’.

The question of how we go about locating those ‘that’s which are relevant
to this rule, will involve us in considering the character of all of these rules
when they’re operating across utterances. And let’s do that, briefly, anyway.
The occurrence of these rules, being not located within a single utterance,
poses for us a problem: What form are these rules supposed to have, insofar
as they are somehow regulating or guiding the Members’ actions; and what
kind of procedures for analysis do they pose?

At least a rough version of the problem might be as follows: The fact that
there are these rules, and that they have a clear locus within a single sentence,
say, operates to provide a maxim which is what we can call a Second Speaker’s
Maxim, though ‘second speaker’ has to be given what may be seen as a special
formulation. Somebody who wants to make themselves a second speaker to
some utterance can have as a maxim for him that if he wants to tie his
utterance to that of a preceding speaker’s, he can formulate some part of his
utterance as a second part of one of these rules, if he can use some second part
which can be tied to some then-made ‘first part’ of a then-made ‘first
speaker’s’ utterance. That is to say, somebody who wants to make
themselves a second speaker, can make somebody a first speaker. It can be
anywhere in the conversation. He makes some part of what some speaker
did, a ‘first part’ for one of these rules. He doesn’t start out, we take it,
with a set of marked objects: First Parts for these rules. But by his speech
he makes some part of what some speaker did, a ‘first part’ for one of the
rules.

152

Part II

So the analysis starts out with the locations of the second parts of these
rules. And it’s in that sense that someone formulates himself as a ‘second
speaker,’ i.e., the maxim doesn’t govern first speaker’s actions. What we’ve
got then, is something like this: There are these rules which may have an
unspecified environment of use, which could be a single sentence, some
sequence of sentences by a single utterer, or the utterances of several persons.
If they’re going to be used by some group of speakers, then it’s the business
of somebody who is formulating himself as a second speaker to invoke them
and provide for the relevance of a first item by using something that is clearly
a second item. When someone does such a thing we get, quite obviously, a
sense in which he ties his utterance to that of some preceding speaker.

Let me indicate that it’s perfectly possible to set oneself up as a first speaker
by the use of other sorts of rules. For example, if you ask a question, then the
fact of a question sets one’s position as ‘first speaker’ to which the rule Q — *
A, Question Followed by Answer, is locatable by the fact that there’s an
observable ‘first occurrence.’ And I think it is the case that there are a variety
of ways that persons can formulate themselves as first speakers for even some
of these Second Speaker Rules, though that’s a rather more subtle problem
and we’ll deal with it as we come to handle data in which it’s present.

And I want to introduce an unspecified caveat to the Second Speaker’s
Maxim: It is the case that persons can in one way or another move to prevent
the use of their utterance as something to which another can tie an utterance,
and they can move to encourage it. So while the maxim we set up can operate
independently of what the formulated ‘first speaker’ does, there are things a
speaker can do to effect what treatment their utterance may get. That there
are, is something that is for at least a certain kind of conversation, quite
important.

But for utterances handled by the Second Speaker Rules, we want to start
out with the second item and find a way of finding the first. And we have to
have some clear way of deciding that we have a second item. Some of them
may be quite clear as, at least, possible second items. The various pro-terms
might be quite clear, but we take it that ‘that’ is not clear. How then do we
find out whether we have a ‘that’ which ought to be seen as a possible second
item, so as to start the search for a first, so as to see, then, whether that person
has used the rules to tie his piece of talk to a preceding piece of talk? What
we’re considering then, is the sequence that the analysis has to go through,
and what is has to include.

Apparently what it has to include is an analysis of the syntax of the
sentence. And that’s important. It’s in the first instance, I think, the syntax of
the sentence that tells you whether you have such a ‘that’ as is a likely second
item. The differences of the ‘that’s are the different syntactic functions that
they serve. Those that we’re dealing with here as second items are probably
easier seen as the objects of the sentences, and those that are not are things
like, particularly, conjunctions. Now those sentences in which the relevant
‘that’s occur, have them in a very clear position. “That’s the problem with
society.’’ And what’s invoked by “that’’ is the predicate of the preceding

Lecture 4 153

utterance; that is, the problem of society is that if you go out and hotrod, etc.,
certain things will happen to you.

And for “I decided that years ago,” ‘I decided years ago’ is an adverbial
phrase, I suppose; ‘that,’ the object, replaced with ‘I’m a chicken shit.’ And
I take it that that at least holds for the material that we have. I also take it
that you oughtn’t to get terribly frightened, since you could easily enough see
what the object of the sentence is, as compared to ‘that’ being used as a
conjunction in, for example, ‘‘I still say though that if you take ...”

But the important thing is that to locate those second items we have to go
through some crude or not syntactic considerations, and it’s not the word’s
occurrence that does that for us. If we then locate a ‘that’ that is an object of
a sentence, we have the basis for then proceeding to consider whether there is,
and what is, that thing to which it’s tied. (In the materials we have, they tend
to be whole predicates, I guess, of preceding utterances.) and if not whole
predicates, then pretty near whole predicates.) Now that’s something that we
haven’t previously considered. That is, we have not previously found
situations where syntactical analysis was required before we could do the
analysis we needed to do.

There are a variety of other things which are, in the case of this
conversation, perfectly simple to see. Such things as the simple use in some
second utterance of the same subject or the same subject plus verb or the same
verb, as is used in a previous utterance. There are a bunch of such things; for
example, the use of ‘you’ in Joe’s ‘‘if you take a big fancy car out on the road
and you’re hotroddin’ around you’re bound to get caught,” followed by Mel’s
“You wanna try and do it right so you do not get caught.” That ‘you’ is an
exceptionally important phenomenon. It’s a definite pronoun, and it has no
special reference at all. It certainly doesn’t mean you, the listener. I’ll consider
that sort of ‘you’ when I get finished dealing with these rather surface aspects
of the way the conversation is tied.

Verb phrases recur in first and second utterances; there’s “picked up”

. . .“picked up” in some sequence, ‘tear down”. . .“tear down ” in another, 1
and others.

Now there are some uses of the rules which start out by having the whole

1 The reference here is probably to the following:

Joe : and the guy will pick the guy up in the dirty grubby tee shirt before I

before he’ll pick me up. Just- just for-

Mel : Not many people get picked up in a Pontiac station wagon.

and

Henry : Hey don’t tear him down

Joe : I’ve been torn down for-

Mel : Ok

Henry : We got company

154

Part II

rule introduced in a single utterance, and then, by virtue of the fact that the
rules are repetitively operatable - as we said earlier, you could have N ■*— P
•*— P, indefinitely - what we get is: N •*— P in one utterance, and then the
continuation in the second. So we have, “Joe, face it you’re a poor little rich
kid,” then several lines down there’s an elaborate way of relocating that
subject, and a continuation of the use of ‘you’ as the subject thereafter, quite
clearly tying that to the first use.

Henry : Joe, face it, You’re a poor little rich kid.

Mel : Okay Now you’ve got that out of your system. Now you’re a

poor little rich kid we’ve told you that.

Henry : And we also decided you’re a chicken shit.

Mel : Now let’s see what else can we decide about you?

I’ll consider this one, also, in more detail, because it’s not a perfectly simple
use.

There’s a rather interesting one involved in the following sequence. We get
“Bonneville” •*— “it,” i.e., N ■*— P, and then “Pontiac station wagon”
“it.” They’re talking about the same car, and it’s at least plausible that the
reference of ‘Pontiac station wagon’ is to ‘it’ in the preceding utterance, as its
antecedent.

Joe : In that Bonneville of mine? I could take it out with me . . . and the

guy will pick the guy up in the dirty grubby tee shirt before I before
he’ll pick me up. Just- just for—

Mel : Not many people get picked up in a Pontiac station wagon.

Joe : Now I agree it looks like a daddy’s- It looks like a damn

mommy’s car.

I said it’s interesting because there’s a tremendous economy in this conver-
sation, and in many such conversations, involving the use of nouns - they’re
just not much used. And when you get one used, it’s doing some work. The
compactness of these conversations is handled in part via the tremendous use
of these pro-terms, which just make up the bulk of talk. That can be seen at
first glance by virtue of the fact that there’s only one use of a personal name
in the conversation, and that use has a major significance, it marks a big
change in topic.

In the case of “Bonneville”. . .“it,” “Pontiac station wagon”. . .“it,” the
renaming of the car is clearly a recharacterization of the car. I say ‘clearly,’ in
the sense that if you knew it was the same car, then you’d take it that
something different is being pointed out about it. The person who originally
used the term ‘Bonneville’ acknowledges that something different has been
done (Joe, with his “Now I agree it looks like a . . . damn mommy’s car”),

Lecture 4

155

and some people who have seen the conversation and didn’t know it was the
same car, took it that in fact an attack was being made, without regard to
whether ‘Bonneville’ and ‘Pontiac station wagon’ were different names of the
same object. The fact that that’s done by change in name would suggest that
whatever generic critique one could make of Russell’s notion that names are
disguised descriptions, nonetheless in this sort of situation that’s the way the
names are being used. So that ‘Pontiac station wagon’ is a transformation of
the name of the object being initially referred to, which transforms the
character of that object from something which, let’s say roughly at this point,
could be a thing which you could approximate hotrodding in, to something
which you couldn’t possibly approximate hotrodding in. Why you couldn’t
possibly, is something we’ll consider eventually.

Let me just note about the range of sequencing rules, that they can be used
just as sheer indicators that there’s a conversation going on. And one way you
can characterize people who can’t converse is that, so far as you can tell, you
could perfectly well disorder the conversation without having any effect on it.
That is, you could juxtapose any of the parts. Given the way these things
operate, if we have these rules tying pieces of conversation to other pieces, you
can’t disorder them. You can’t put an utterance that contains the second for
some particular first, ahead of it. It matters what sequence they come out in.
And people make it their business to make it matter.

There are also some much more interesting, perhaps less obvious, and in
any case less overt and ‘technical’ - in the sense of having to do with the
names of these words - ways in which this conversation is tied together. But
to get at them I’ll have to do a whole big analysis and I want to deal with
things that are right on the face of the conversation before we move on, so you
can see first how obviously locked together it is. Therefore, I’m going to move
to some other surface features of the conversation before dealing with the
more elaborate ways it’s tied together. In doing that. I’ll be doing something
which I ought to have done previously, and which may give some sop to
people who are interested in personality . 2 And that is, there are some really
quite striking ways that, even in this one page, these three guys speak
differently. And these are quite obvious and observable things, though some
of their technical features may not be obvious. For example, this fellow Joe
recurrently uses a pattern like “you’re you’re bound to get you’re bound to get
caught.’’ There are three or four such uses. It’s not an unheard-of thing. And
it’s important to see that it’s not an unheard-of thing; that it is something that
has been observed, and observed as characteristic of something. This is not to
say that if it is characteristic of something, then he is one of them. But it is
characteristic of the speech of very young children. It’s something called
‘buildups.’ For young children, it may have as its basis the fact that they have
a very limited grasp of grammar, and they literally assemble a sentence by
putting its parts together, piece by piece. If you look at a book I’ve mentioned
before, Language in the Crib by Ruth Weir, she has a discussion of these

2 The title of the course was ‘Culture and Personality.’

156

Part II

things (pages 81-82), and a bunch of examples. The kid there is, I think,
something around two. But any of you who have heard young children talk,
will have heard that sort of thing. It’s obviously not simply a stutter. And it’s
extremely characteristic of this fellow, Joe.

Another thing that is sort of striking occurs twice in Mel’s speech, and I
don’t think anywhere else. It’s not a unique usage at all, and it’s perfectly well
examinable as to what it can do. And that is, he uses the syntax of questions
without making - for utterance-rule purposes - questions. That is, he builds
a syntactical question without intending that at the end of it he would stop
talking and somebody would answer. He does that in quite obvious ways by
using the reversal pattern of questions: “Now did you do it right,” for
.example, and “Now let’s see what else can we decide about you.” That is,
“. . . can we . . . ,” “. . . did you . . .” are such kinds of things. And I’ll
consider those, also.

The third fellow, Henry, has what looks like an exceptionally idiomatized
way of talking. Three of his whole utterances are single pre-constructed
sentences, akin to proverbs in a way. That is, they are sentences that he didn’t
have to construct for this occasion, that can by used by whomsoever, that
come on just as that - a sentence. Things like, “That’s the problem with
society,” “face the music,” and “we got company.”

Lecture 5

Tying rules; Insult sequences

I’ve been examining how the parts of this segment of conversation are tied
together, dealing first with some very small ‘pair-ties.’ We’ll move on to
much more elaborate ones as we go along, and then get into what this
conversation is about, and how it works with respect to that kind of matter.
The kinds of rules I’ve been offering are by and large not at all constrained by
what the conversation happens to be dealing with. Eventually we’ll see that
there are no untied utterances in the segment. And by that I mean, no
utterance for which there is not at least a single tie. We’ll find that for some
there are far more than a single tie. I haven’t worked out a measuring system
to compare conversations in this regard, but I take it that that might be
doable in any event.

Let me go over some of the rules I’ve already given. We have, for example,
some purely-internal-to-a-sentence rules, like the one for making an indepen-
dent clause out of what might have been a sentence, and for making your own
utterance a dependent clause in that sentence, and for making your own
utterance a dependent clause in that sentence. It runs something to the effect
of: If some given utterance is ended by a noun, then participialize that noun
and use it to begin your own utterance. And of course, besides participializing
that noun you have to be otherwise syntactically consistent with the preceding
utterance. Just consider: “We were having dinner”. . .“dining on roast beef.”
You can make a whole bunch of them in just that way, although that’s not
the only way you make dependent clauses.

The second rule concerns prepositional phrases. They consist of a preposi-
tion plus a noun phrase. If an utterance ends with a preposition, then if you
begin the next with a noun phrase that could be the noun phrase for that
preposition, you tie your utterance to it.

A third rule concerns the use of conjunctions. Now the conjunction is rather
more complicated because the syntactical consistency is much more impor-
tant. If an utterance ends with what might be a sentence, if you begin the next
with a conjunction and otherwise make it syntactically consistent, then again
you can tie the second to the first.

Then there are a variety of rules which, while they could operate within a
sentence or within a single utterance composed of several sentences, they can
also operate across utterances. Given that for these that I’m going to list, we
start out with the second item for the rule and not the first, it might be
convenient not to write them as, e.g., N P (N followed by P), but as P — ►
N, so that we know that we’re starting with P. So there’s that one for
pronouns and nouns, there’s the same sort for pro-verbs and verbs, there’s the

157

158

Part II

one for ‘that’ (with a special constraint on syntax) and predicates; now there
are some more.

Some of the others, once we mention them, will permit us to see a rather
clear difference between those things which we identify in the first instance
from what turns out to be the second object, and those we identify by noticing
the first object. For example, we can look at a question-answer sequence,
which is as clear a version as you want of starting out with the first part. In
this conversation you get things like: “What school did you go to?’’ I went to
Palisades.’’ and “You’ve been kicked out of school once.” “Yeah.” Now for
those, it’s by virtue of the fact that the first item provides that the next is
going to be a second, that those seconds are clearly seconds. They are not
otherwise so identifiable. They’re not even identifiable as answers to questions.
But there’s no doubt, I take it, for anybody, that things like “Yeah” in that
sequence, are second items. But what we have in these other kinds of rules,
is that the first items are not identifiable as ‘first items.’ It’s by virtue of the
fact that the second is clearly ‘a second to something’ that we go back and find
a first, as compared to, say, the question-answer sequence where we identify
the first. That means, for one, that you get a kind of distribution of
compactness in the question-answer sequences. Given the fact of the question
providing the relevant sense of the thing that ought to follow it, nothing
much has to go into the answer. And a zillion questions could be answered by
‘yes’ or ‘no.’

Now we can find some instances of this question-answer sort of rule in this
piece of conversation. There is at least one command and then return-to-
command - it could be either acceptance or rejection of the command, but it
would be clearly seen as one or the other. For example, ‘Joe, face it . . . ,”
“Yes . . .” (I’m not saying the ‘yes’ is an acceptance. I’ll analyze it eventually.
It’s a much more careful thing than an acceptance in that case.) It is by virtue
of the command being seen as the first item, providing for a relevant second,
that the seconds have their character, though some answers are obviously
answers - not many.

There’s one which you can play conservative or non-conservative on, as to
how you write the rule, but let’s say: Incomplete Utterance followed by
Interrupter Term. I say you can play this conservative because you might not
take it that the utterance was clearly incomplete. And you might start out
with the fact that it looks like you’ve got an interrupter term, and work
backwards: interrupter term — ► incomplete utterance.

I take it that that holds in its strong version - that is, incomplete utterance
first - for the sequence:

Mel : Now let’s see what else can we decide about you?

Henry : Hey don’t tear him down

‘Hey’ is clearly an interrupter term. You might not take it that it’s obvious
that “Now let’s see what else can we decide about you” is an incomplete
utterance. To get some feel for that, reconsider the earlier statement that is

Lecture 5

159

syntactically alike to it, “Now did you do it right.” And I want to cross out
the question mark at the end of “Now let’s see what else can we decide about
you?” because in my last listening to that piece of conversation I find that I
now think that there is no intonation at the end that leads one to feel it’s a
question, though of course it has that reversal construction of a question, i.e.,
not “let’s see what else we can decide ...” but “let’s see what else can we
decide ...”

There is a class of tying verbs that are used. There are a lot of verbs that
may tie, but there are some verbs which, when used, are just tying verbs. Like
‘agree.’ When somebody says “I agree,” they’re tying that statement to some
other statement of somebody else’s. So you could write it as some utterance
•«— tying verb, but it would probably be done as tying verb — *• some
utterance.

Now there are some things which can occur within a single utterance or
could even occur in single sentences, which also can occur across utterances.
When they occur across utterances they are clearly serving to tie utterances
together. And those are what we could call lister terms. The most obvious
lister terms are, of course, things like the ordinal numbers: First, second, third,
as we get in the next segment of this conversation, in the series starting with
“Well first of all you must be crazy . . .” But there are some used in this
second segment to do that kind of work, that are not the ordinal numbers. A
series can be explicitly started with a first, L, — ► L 2 . Or you can have a first
occurrence of a lister that is making itself follow something else, where the
other has not been specifically introduced as the first of a list. In this segment
Mel says “Now you’re a poor little rich kid, we’ve told you that.” You might
say that ‘now’ is a lister term. It’s not clearly one, it has plenty of other uses,
and you’d have to make a good argument for it. But it’s followed by “And
we also decided ...” which turns it into a ‘first’ where it could have been just
a single.

A lot of other rules can be constructed, but with these you can certainly get
some idea that these people were in some way or other talking to each other,
and making the fact that they were talking to each other a big part of their
business.

Now, what I’m going to be doing is taking small parts of a thing and
building out from them, because small parts can be identified and worked on
without regard to the larger thing they’re part of. And they can work in a
variety of larger parts than the one they happen to be working in. I don’t do
that just as a matter of simplicity, but as I mentioned earlier in the course, the
image I have is of this machinery, where you would have some standardized
gadget that you can stick in here and there and that can work in a variety of
different machines. And you go through the warehouse picking them up to
build some given thing you want to build. So these smaller components are
first to be identified because they are components perhaps for lots of other
tasks than the one they’re used in. But then they’ll be fitted together into some
actual single larger component in this case.

Okay, let’s look at this one for a while:

160

Part II

Henry : Joe, face it, You’re a poor little rich kid.

Joe : Yes Mommy.

Henry : Face the music

We could start off simply enough by saying that it is a recurring part of, let’s
say at least, teenager’s conversations, that an insult is properly followed by
another, or a return to it.

In fact, that little thing itself is the basic component of some very
standardized teenagers’ games. They are alternation games, they have names,
and one of the names is in fact used later on in the conversation. There is a
considerable literature which pretty much first emerged into literature in
a paper by John Dollard in 1957 called ‘On playing the dozens,’ which is a
Negro kids’ game, and it’s an insult game. The kids engage sequentially in
insulting each other. They tend to do it before some audience, where the
audience, by its reaction, decides that the game is over. In that form it’s
been well written-up in a variety of papers. Dollard’s paper is in a book by
Roger Abraham on that sort of phenomenon, called Down in the Jungle
( 1964 ).

The game was apparently not much thought to be one that white high
school kids play, but they do. And one of the names in that environment is
‘tearing down,’ which is what the action eventually gets called in this
sequence. It is not, however, to be seen as getting its analysis by some simple
historical expansion - which is one way that people have considered it - that
is, that it passed from Negro kids to white kids at some time. Because some
of its features, like this ‘Mommy’ bit, are fabulously old.

The ‘Mommy’ return is extremely classic in such kinds of things. I’ll give
you a quotation. This guy Eric Partridge, who writes all these things up, has
a book called The Shaggy Dog Story, and in it he reports what he calls an
ancient Greek story.

A pert youth meeting an old woman driving a herd, called “Good

morning mother of asses.” “Good morning, my son,” she returned.

Apparently that’s one of the classic ways that you handle an insult, i.e., if it’s
an insult for which the most elegant return would be to make my status a
consequence of yours. And kinship obviously is the most powerful way of
doing that, like “You’re an ass.” “Thanks Dad.” Those kind of things are
obviously extremely powerful and by no means new, by no means to be
accounted for by the fact that in Detroit the racial barriers are broken down
so that Negro kids and white kids are closer to each other - except
incidentally.

There are several reasons I’m going through the fact of this game. One is
that by treating the first event as possibly the first event in such a game, by
doing a second, that’s a way of giving a special characterization to the first
insult as something that is not serious and is not going to be heard as serious.
It also sets up a challenge, which is, Okay you started, I came back, go ahead

Lecture 5

161

if you’re willing to. And Henry withdraws. That is, the move that would
accept the sequence as a game and continue it, would be for him, since he has
a slot, to come back with another insult. But he closes it off with “Face the
music.’’ It’s quite apparent that what he’s done is to withdraw, in that when
it rapidly turns out that he has an ally in that thing, it starts up again, as you
can easily enough see.

Henry : Joe, face it. You’re a poor little rich kid.

Joe : Yes Mommy.

Henry : Face the music

Mel : Ok Now you’ve got that out of your system. Now you’re

a poor little rich kid we’ve told you that.

Henry : And we also decided you’re a chicken shit.

While we’re on this sequence, we might as well note that one of the things
we earlier pointed out is also present here. I proposed that if it were the case
that we had the form of pronoun-noun relationship where the pronoun
actually precedes the noun and is not the second that’s discovered first, but is
first, then something special is going on. And the ‘it’ in “Joe, face it’’ is such
a one. One pretty clear thing it’s doing is, it keeps that utterance open to then
flip out the item that completes the N-P pair.

There’s a whole bunch of ways that statements that might be closed by
virtue of the fact that they’re sentences, are kept open in this first segment. I
pointed to some of them already, though I haven’t worked out how they
work. For example, those special kinds of questions that this fellow Mel uses.
Another is that even if the game itself doesn’t go on, and one is restricting
oneself to an insult-and-retort sequence as one’s expectation, then apparently
one of the things about the insult-retort sequence (and this is another
tradition which is perhaps independent of or pre-existent to this game) is
that there’s a big crucial thing on having the last word, for some reason or
another. And by first producing “Face it’’ and saving the formula “Face the
music,’’ then, whatever the retort is, Henry has the last word, because the
formula is tied to his first usage, and he can then fit it in and close the sequence
off. So, where you have a formula that you’re invoking by splitting its usage,
that may be a significant thing to do, and in this case it clearly is. And
there’s plenty of literature about having the last word. There’s a classic
thing to look at, an absolutely fabulous book I mentioned before, The Lore
and Language of School Children by Peter and Iona Opie (Oxford University
Press).

Now I want to deal with another little one, one which what I have to say
is somewhat beyond my level of believability - which I guess is much more
extravagant than other people’s - but nonetheless it’s possible. And that’s the
sequence:

Henry : And we also decided you’re a chicken shit.

Joe : I decided that years ago. Hell with you

162

Part II

The question is, what kind of a return is “I decided that years ago, hell with
you.” Clearly ‘‘hell with you” is of some relevance. It’s an indication of
courage, or something. Now it’s perfectly clear that “I decided that years
ago” is not an agreement. It’s instead, a rejection. Why should it be a
rejection? It looks like a perfectly good agreement, nonetheless it’s perfectly
clearly not an agreement but a rejection. There is an obvious enough
formulation, except it’s as I say, beyond my belief that they could be that
ingenious, and that is that what he’s doing is to set up an antinomy. And he’s
using the fact that he’s setting up an antinomy to undercut the statement. An
antinomy is a statement that, if it’s true, it’s false, and if it’s false, it’s true.
The classical one is “I’m a liar. Everything I say is false.” Here, if it takes
courage to recognize that you’re a coward, then to have recognized that you’re
a coward took courage, and therefore you couldn’t be a coward.

Given such a possibility, I would then try to see, is it the case that people
in their ordinary affairs do use antinomies. And I haven’t really looked very
hard so far, but I found at least one, and it’s a classic one, also. It comes from
this book, Style in Language, edited by Thomas Sebeok, a paper called ‘Oral
styles of American folk narrators’ by this American folklorist, Dorsen, on
page 41. He’s reporting some story he was told:

The next supposedly true happening, where Art is asked to tell a lie and
says he has no time because so-and-so has just had an accident and he
must go to a doctor - which is a lie - is an international folk tale
attached to various American yarnspinners

Lecture 6

‘You'

I’ve been talking about this bunch of tying rules. There are a lot more, even
of the sort that we’ve been laying out, apart from any that are more intricate
than those, but that’s enough for now. The question is, so there they are, why
bother talking about them? What is it that they do? To some extent it’s
terribly simple to see. It’s one sort of task for the hearer of some statement to
determine what it is that the person who produces some noun is referring to,
if he’s referring to anything: What’s an ‘automobile discussion’? What’s a
‘car’? etc.; it must mean this or that. That’s one sort of task. When the tying
rules are used, another sort of task is imposed on the hearer: To decide what
it is that the tied term - for example, ‘it’ - refers to, requires finding
somewhere in the conversation that the term it ties to occurs.

So in the use of these, there’s then a required piece of work for the hearer,
which involves collaboration in making out the conversation. Understanding
the term implicates the hearer in the conversation - and of course provides
that in the very use of such an object the speaker has implicated himself in the
course of the conversation. The use of these, then, is not to be seen as simply
a way of, for example, avoiding redundancy or making variety or whatever
else, but it provides an order of work - and is produced by an order of
work - which is at least to some extent different than the work involved in
using a name, etc. So that, at least in the first instance, is what the use of these
tying rules involves. And that’s then, the way in which they make for the
sense that the various participants have, that a conversation is taking place;
where the participants are implicated, and they are involved in working at all
of its parts, as a set of parts.

Now the work that’s involved can be sort of minor in that someone may
claim, or may in fact not be able to see what it is that some pro-term refers
to, i.e., what term it refers to. But it can also be extremely important. It can
be the case that some usages provide that if their hearer wants to play dumb,
then he’s going to have an extremely touchy task to do that. That is, the
routine use of these things may operate to provide a great deal of information
for the one who uses them; for example, that on some occasion the other does
in fact understand what one is talking about, by virtue of the fact that he can
continue to put in terms where, quote, nothing explicit has been said - or not
much explicit has been said. Eventually I’ll give you some instances of usages
which are strategically very powerful in the sense of forcing someone to
acknowledge that they do in fact understand something, where the fact that
they do understand it has tremendous consequences. But for now, I just
wanted to raise it. The initial remarks about tying rules would be enough to

163

164

Part II

give them some strong warrant. The things I will mention eventually are just
elaborate technical developments on them, and they’re of some interest but
they’re not that enormously core.

Let’s go back, then, to dealing with bits and pieces. In the first pair of
utterances after the introductory sentence of the second segment, we have
‘you’ used as basically the subject, i.e., “. . .if you take a big fancy car out on
the road and you’re hotrodding around you’re bound to get caught and
you’re bound to get shafted.” “Now did you do it right. That’s the challenge.
You want to try and do it right so you do not get caught.” I want to make
some remarks about ‘you.’

First let me list some references to the extremely large literature on terms
like ‘you;’ terms such as I, you, he, this, here, now, it, and the like. They’ve
been subject to a lot of study by philosophers and logicians, and if one is
interested in conversation, one might well look at that literature to see what’s
been done on the phenomenon of conversation. Some of the literature is at
least historically interesting. There’s chapter 7 of Russell’s book, Inquiry into
the Meaning of Truth, and he calls them ‘egocentric particulars.’ Then there’s
Nelson Goodman’s book, The Structure of Appearance, chapter 11. He calls
them ‘indicator terms.’ Then there’s Reichenbach’s book called Elements of
Symbolic Logic. He has a long chapter on the analysis of conversational
language, which isn’t what it says it is; not at all. It’s important to see that it’s
not - and to see that nonetheless it’s pretty unique in saying that it is, anyway.
That’s chapter 7, and he calls them ‘reflexive terms.’ Then there’s Quine,
Word and Object, section 21. He also calls them ‘indicator terms.’ Then there’s
a lot of articles in the philosophical journals. I’m not going to cite them all.
Some of them use the names I’ve given you, but one that’s relevant in that
it uses its own name is a paper by Bar Hillel in Mind, 63 (1954), titled ‘On
indexical expression.’ That’s what he calls it. Then there’s an old monograph
by a linguist named Cullenson, called ‘Indicators,’ Language Monographs, 17
(1937).

The core thing about these terms, and why they’re of interest to the
logicians, is that they have an extraordinary transiency of reference. So the
issue is, if you’re dealing with ‘the truth or falsity of propositions,’ then how
are you to interpret a proposition which has the term ‘now’ in it, or ‘he’ in it,
or ‘you’ in it, etc. Much of the work has been directed to trying to find a way
of rewriting statements or sentences that have those terms in them, so as to
provide for the fact that they could be true or false. I won’t be considering
such issues.

But there’s one immediate consequence for my previous discussion. One of
the things you ought to see is that if any such terms occur in immediately
juxtaposed utterances, that would not be necessarily at all an instance of
something that you would want to call a tying-rule case. My proposal that ‘If
the same subject appears in two consecutive utterances then those utterances
are tied,’ would not at all obviously be the case if ‘you,’ and then ‘you’
appeared. For one, if there are two different speakers, they’re not at all
referring in any obvious way to the same person. In fact, if you’re going to use

Lecture 6

165

these in focusing on the tying operation, what you would want is that
whichever term appears in the first utterance, the alternative term appears in
the second. You could have an instance of tying if you had ‘y° u ’ ‘I,’ or T

— *• ‘you,’ where ‘you’ — ► ‘you’ is not at all an obvious such case. So these
terms pose a problem for working on the tying operation, but it may be
worked out. In any event, the usual ‘you’ — *■ ‘you’ problem is not present in
the example we’ll be considering.

In this case, “. . . if you’re hotrodding you’re bound to get caught . . . ,”
“You want to try and do it right so you do not get caught,’’ we have what
is a very recurrent use of ‘you,’ and it’s what, if Reichenbach were to talk
about it, he would immediately translate as ‘one,’ ‘someone,’ or something
like that. It is not referring to the person being addressed. It’s what we could
call, to start off with, an ‘indefinite use.’ And one of the things we want to ask
is, how does it come about that ‘you’ is used to make that indefinite
statement? Why don’t they say ‘one’? Are they equal?

I think there are more than a couple of things involved, and I’ll try to
mention most of them that I know of. Let me first discriminate a couple of
possible senses of the term ‘ambiguity.’ One characteristic way it would be
used is in an either-or relationship - either you mean this or you mean that,
it’s not clear which, and they don’t have anything much to do with each other.
Another usage of the term - and that usage is most considered when you’re
dealing with poetic language - is this-and-that. For poetic language, one
question is, how do you go about building richness into a compact form. One
way is to get words which mean this and that, whatever this and that may
mean. And the usage of ‘you’ that we’ll be considering is, I think, a case of
the latter. ‘You’ is a very good term for attempting to build ambiguity of the
this-and-that sort.

I think the analysis ought to run something like this: What we have to do
is to try to construct what a procedure might be for determining what it is
that’s being referred to when somebody says ‘you.’ Is there a procedure? If
there is one, what does it look like? And what are the consequences of there
being one?

Now ‘you’ in English is, at least in the first instance, systematically
ambiguous, in that it does not discriminate between singular and plural
usage. And in the ambiguity of ‘you,’ one has the this-and-that format; that
is to say, it is ‘you’ (you alone) or ‘you’ (you and others). When a person hears
‘you,’ they then go through a procedure of deciding what it refers to. And
that procedure has a first step, where if the first step seems fully adequate, that
ends the procedure, but if it’s not, then one goes on. The first step in the
procedure is the consideration of the applicability of the singular use, and if
it’s the case that one comes up with, well maybe that, but certainly not only
that, then one moves to the second step and engages in finding, now, some
plurality for which the ‘you’ would be correct. And that plurality could be, for
example as I mentioned earlier, some category which, once it’s found it would
turn out that the singular is a member of. It’s also the case that one of the nice
things about the non-differentiation and the plural character of you’ is that

166

Part II

there is no limitation essentially on the size of the population it can be
referring to.

The singular ‘you’ is used as simply a way of dealing with the persons
present and being spoken to in an interaction. Plural ‘you’ is no longer
controlled by that fact. It’s not at all necessarily the case that it’s only referring
to the multiplicity of persons present and being spoken to. Indeed, as I
mentioned earlier, ‘you,’ even in its singular case, is regularly a way of
referring to that member of ‘they’ who happens to be present. “You
Negroes.’’

That openness of the plural ‘you’ means that ‘you’ can in fact be a way of
talking about everybody’ - and indeed, incidentally, of ‘me.’ But there’s at
least a tactical difference between saying ‘we’ with that intention and ‘you’
with that intention. The difference is, ‘we’ has some plural reference,
automatically. And it does not have, automatically, a plural reference which
includes the person being spoken to. This is in English. In some other
languages it’s the case that ‘we’ can only be properly used if it refers to [me
and the person I’m talking to}. So you do not say ‘we’ when you’re referring
to [me and my wife, and not you}. In English, we do. And ‘we’ can
then - intentionally by the person speaking, or by the decision of the person
being spoken to - exclude the person being spoken to. In contrast, if you use
‘you,’ it at least includes the one you’re speaking to, and on their option or on
your intention, insofar as those coincide, it can refer to anybody else, or to
some category which includes everybody else.

And those differences are extremely carefully focused on by speakers. I
have a lot of very subtle usages which turn on those differences; for example,
a woman is asked “Why do you want to kill yourself?” and she says “Well,
you just want to see if anybody cares.” Now that use of ‘you’ in this case
surely refers to her, but refers to her as a member of ‘anybody,’ and thereby
provides that it is only incidentally her reason, but it’s anybody’s reason, and
thereby is not attackable as peculiar. It is offered as proverbially correct.

Those kinds of uses are recurrent, and apparently quite powerful. I have
one I just pulled out of the paper the other day, involving an interview
between vice-president Humphrey and a Japanese newspaperman, on some
relationship between Japan and the United States over control of atomic
weapons, I think it was. And Humphrey is asked about America’s policy, and
whether Japan would have more say. And, although he’s talking about what
‘we’ will do, he formulates his remarks in terms of ‘you,’ e.g., “You want to
give other people a chance,” etc. This involves now putting the Japanese who
read it in the position of assessing his problem as though it were anyone’s
problem in such a situation, and it’s nothing peculiar to America’s position
that involves them in being hesitant; anybody in such a situation would be
hesitant.

Now, one of the core things about this use of ‘you’ is that if in the first
instance you find that the singular ‘you’ is inappropriate, you do not move to
a sense of it which excludes yourself, but to a sense of it which is much larger
but at least includes yourself. It may also, by interpretation, include the

Lecture 6

167

person who has spoken - as you see it, or as they see it, or obviously. But
nonetheless, you are also included, which ‘one’ does not do. And ‘we’ does not
do, and any of the other terms do not do - ‘he’ doesn’t, ‘they’ doesn’t, etc.
But ‘you’ uniquely does. So it isn’t the case that ‘you’ and ‘one’ are
equivalent. 1 And it does seem to be the case that the procedures for
determining what it is that ‘you’ refers to — if the procedure is something like
I make it out - is crucial to the special use the ‘you’ gets and that no other
pronoun term gets. That is, ‘you’ is the term which gets used in specifying a
proverb, or a proverbial type of frame.

Returning to this sequence, there are other things involved in the use of
‘you.’ One of them is another kind of ambiguity. We get this statement of
Joe’s, “if you go hotrodding you’re bound to get caught.’’ If it’s some generic
type of argument of an if-then sort (though there’s no explicit ‘then’), it could
be delivered as, “if anyone goes hotrodding . . . ,” “if a kid goes hotrod-
ding ...,’’ etc. What the ‘you’ does is, beside making it an argument, it
borrows the form of a warning. That’s the way warnings get delivered. “If
you do the following, then this will happen to you.’’ And the relationship, for
kids anyway, of those generic if-then types with warnings, is probably
something quite ambiguous. They first learn the warning version. They then
have to learn to separate the pure argument type. And for some matters, I
take it, it must be quite important that one never makes a complete break,
and that there are ways of making the assertion not merely an argument, but
also a warning, or a piece of advice. And clearly one way that would be done
is to use the form of the warning, which involves the use of ‘you’ in that
phrase. More generally, if you count the number of uses of ‘you’ in these
proverbs, and their power - which is tremendous - you have a lot of
material.

Now I want to consider the sequence:

Me/\ Ok Now you’ve got that out of your system. Now you’re a
poor little rich kid we’ve told you that.

Henry : And we also decided you’re a chicken shit.

When we started out the discussion of this segment, I pointed out right at the
beginning that we have here, again, as we have in the first sentence of the
segment, a collaborative production of a sentence. In this case there are several
ways that the parts are tied together. It’s not simply the ‘and,’ but the subject
used in the first utterance is retained in the second, i.e., the ‘we’ - and that’s
extremely crucial. And the relation is further established by the ‘also,’ which
provides that it’s building on what has just been said.

Simply by virtue of the set of properties we got, we might feel quite
confident in saying that what we’re getting at this point is the forming-up of

1 Several people in the class speak up with difficulties about the references of the
terms, including a disagreement with Sacks’ proposals about what ‘one’ does and
doesn’t do in contrast to ‘you.’ Sacks’ response is that “ ‘One’ doesn’t include the self
or the other necessarily. [It] doesn’t include anybody necessarily.’’

168

Part II

a second group within the group, which second group consists of as members
the two participants to this sentence. Let’s look at it a little more - not that
we’re going to find out that it’s wrong, but just to see how it works out. Until
that point, we really haven’t had such a demonstration of the two-against-one
characteristics of this conversation. When Joe got finished with his first long
speech, the two others spoke essentially at the same time and took different
positions.

you’re bound to get caught and you’re bound to get shafted.
We- look

Now did you do it right. That’s the challenge That’s the
challenge You wanna try
That’s the problem with society. Hahhh
and do it right so you do not get caught.

That’s the

And when we got to that point where Henry starts to attack Joe, he did it a
bit on his own, and closed it off - as I’ve suggested anyway.

Henry : Joe, face it. You’re a poor little rich kid.

Joe : Yes Mommy.

Henry : Face the music

Now Mel comes to the rescue, in a way. He ties his own statement back to
the preceding attack on Joe, with the ‘now,’ and the reassertion of Henry’s
statement.

Henry : Face the music.

Mel : Ok Now you’ve got that out of your system. Now you’re a

poor little rich kid we’ve told you that.

He reasserts the statement as ‘we:’ “we’ve told you that.’’ That ‘we’ can have
a strong and a weak usage. It could be strong if it were the case that what he’s
saying is, ‘Henry told you that, and he is our representative.’ That is, the
single person who spoke is a member of ‘we.’ That would be a chancy kind
of thing to do, in two ways: One, Joe could say, “No, ‘we’ didn’t tell me that,
he told me that.’’ And two, Henry could say, “No, ‘we’ didn’t tell him that,

I told him that.’’ So he’s taking kind of a chance there, to go out on his own
to establish what would be protective for the two of those fellows if it gets
accepted. Given the possible chancyness of it, he does in fact establish it was
weakly correct by virtue of the following: ‘We’ can equal, besides ‘he, as a
representative of us,’ also, ‘he plus I.’ And he has in fact said it just now, and
he saves the “we told you that” until he said it, so it’s weakly true at least,
that ‘Henry said it and I said it,’ and that is ‘we.’

Joe :

Mel\

Henry :
■Mel:

■ Henry :

Lecture 7

‘Hotrodders’ as a Revolutionary
Category

At the end of last time I was talking about that statement of Mel’s and most
particularly the use of ‘we’ in it:

Mel: Ok Now you’ve got that out of your system. Now you’re a poor

little rich kid we’ve told you that.

In it, at least as I see it, there’s two possibilities laid out. One is an offer made
by Mel to Henry that the two of us are a group, of which what perhaps either
of us say is said representatively for both. Alternatively, there’s weaker use of
‘we’ which makes his statement at least safely true, i.e., where ‘we’ means in
its weak sense, ‘you said it before and I said it now’, and thereby he needn’t
be subject to any total rebuff. Now Henry takes him up on the strong use, and
accepts the formulation in its strong sense, and in doing so he moves to a
whole range of important doings.

First of all, with his “And we also decided you’re a chicken shit,’’ he tacks
on something that had initially been proposed by him prior to this
eight-minute segment, as something that ‘we’ did, also. Since he was
representative just a moment ago, he was also representative a while ago. In
doing that, he also is involved in more sharply posing the fact that there has
been a change in topic, as well as a change in the formation of the given
group. The change in topic is that Joe is now clearly the topic. And Henry
does that by indicating that a list has been started about Joe, with ‘also’ being
this ‘lister term.’ And in doing so, he provides, then, for Mel what it is that
he could do, which is of course to continue with the list. And that is an
extremely safe way to proceed, given that in continuing he’ll be doing so as
a member of this new subgroup within the group, which is in the first instance
a strong one, since it contains at least a majority. There has been a first
amalgamation within the group, with the three of them collaborating to
produce a single sentence. Now there’s two against one. The two have joined,
and the group has a distinct form. It’s not just a bunch of kids sitting around
with no particular structure to the group.

Once we have this majority situation, we get to what could be an extremely
crucial point in it, which Henry proceeds to use in as strong a way as he likes,
and that is, if there’s a majority in the group, then the majority can be
representative of the group. If there’s a representative of the majority, then the
representative of the majority can be the representative of the whole group,

169

170

Part II

and he who leads the majority is thereby the most powerful man in the group,
in a special way: He’s not now just the most powerful man among the three
but nevertheless subject to attack by either of the other two, but he’s the most
powerful man within the majority, protected by the other in the majority, who
would at least be concerned that their group last, and who might fight out the
issue of who’s going to be the leader in that group, but be concerned to keep
it that at least one of them will be the leader.

What Henry rapidly does is, having got into a position where they’re
clearly putting down Joe, he takes his first opportunity to now put down Mel.

Mel : Now you’re a poor little rich kid we’ve told you that.

Henry : And we also decided you’re a chicken shit.

Joe : I decided that years ago. Hell with you

Mel : Now let’s see what else can we decide about you

Henry : Hey don’t tear him down

Joe : I’ve been torn down for-

Mel : ok

Henry : We got company

Mel : oh ok.

Henry takes the opportunity to put down Mel by giving that command,
“Hey don’t tear him down . . . We got company.’’ It is furthermore much
stronger than just putting down Mel, because it also is, in the first instance,
at least an attack on the new entrant. Bob, and also on a lot more, which I’ll
go into. I lead up to this because what I want to do is focus on this term
‘company,’ having just sketched the shifts that take place very rapidly right
here, with Henry accepting the new formation and then moving to take that
over, and thereby to take over the group itself.

The category ‘company’ is, as I briefly mentioned earlier, rather a special
one. And we ought ultimately to consider it in combination and contrast with
a category like ‘hotrodders.’ Before dealing with the basic way it’s a special
kind of category, I ought perhaps to mention some of its import, which can
be seen without regard to what I’ll be saying. And that is, for them to
formulate him as ‘company’ - and we’ve seen how you can say ‘them’ in a
serious sense - is obviously for them to propose that whatever has been done
by whatever institution that might tell this fellow Bob, or his family, that he
has troubles, and whatever way this clinic that they go to might go about
selecting personnel, and whatever position the therapist might be supposed to
have in controlling the group, nevertheless he does not decide, and all those
others do not decide, that Bob is a member. That’s for these kids to decide.
For them, he’s company. And they seem to be able to do that. There is, then,
some kind of independence in their determination of who’s a member, and if
they’re controlling the conversation, then that determination on their part is
indeed important.

Now, holding that, let’s consider what kind of a category this category
‘company’ is. I suppose about the most grandiose way I could choose - which

Lecture 7

171

is the way I choose - to bring the point home, is to quote something which
has become, in its place, extremely famous and extremely fundamental.
Though when you first hear it, unless you know about it, it’s not very likely
that you’ll think there’s anything very special going on. It’s from chapter 14
of Genesis. It goes like this:

A fugitive brought the news to Abraham the Hebrew, who was

camping at the terebinths of Mamre the Amorite, kinsman of Eshkel

and Aner, these being confederates of Abraham.

There’s a phrase there, that is, in the history of biblical criticism, one of the
most fundamental of all in the bible. That is, ‘Abraham the Hebrew.’ There’s
an immense literature on it, and here’s in part what its importance is: It is
among the very few places in the bible that an Israelite is referred to by the
term ‘the Hebrew.’ That term, ‘the Hebrew’ is only used by an Israelite for
self-identification to a foreigner, or by foreigners about Israelites. And what
that’s been taken to mean is, that this section of Genesis was a segment taken
from some document which had not been written by Jews. And the
importance of that is, that it’s the fundamental piece of information which
provides for the possible historicity of Abraham. That is, if there was a
document which was not a part of Israel’s traditions, not written by them, in
which he occurred as a figure, then that’s as strong evidence as they’ve gotten
that it was in fact some person, and not just the name of a tribe, or something
else. (The term ‘Hebrew,’ itself, has other significances; it’s a very generic
term in the ancient Near East, referring to a class of persons.)

The thing is that here’s a category which has become crucial in biblical
criticism because what’s seen about it is that a member would not use it to
refer to himself, except under very special circumstances. It’s a category that
we might say, roughly, is the possession of some group other than its users.
In terms of current large-scale focus, the Muslims have a similar attitude
toward the category ‘Negro.’ And what they want to do, in part, is to legislate
the status of ‘Negro’ into having the same characteristic as ‘Hebrew,’ so that
if you knew somebody said “so-and-so the Negro,’’ you could say about that
somebody, “He’s not one’’ - unless it was a conversation in which the user
was identifying himself to an outsider.

‘Company’ is obviously the same kind of category. It’s used to refer to
some outsider, or it’s used by an outsider to refer to himself when he’s talking
to some member and characterizing himself as an outsider to them. I said it
was a category in alternation to some other category. It is also a category that
is used, by and large, by those who are using the alternating category to
categorize themselves.

We can examine lots of categories to see who is it that owns them. And
we can see that that kind of question can be a really important one. We’ll
see that when we consider what kind of difference there is between a
category like ‘hotrodders’ and one like ‘company,’ and between, let’s say, a
category like ‘hotrodders’ and one like ‘teenagers.’ We’ll also then see some

172

Part II

of the things that people are trying to do when, like the Muslims, they
attempt to propose that Negroes shouldn’t use ‘Negro’ as a category, and
when, like teenagers, they construct and attempt to use a category like
‘hotrodders.’

For them to say that this fellow Bob is ‘company’ will increasingly be seen
as an attack on, not simply the therapist, but in some ways, the ways that the
institutions that get all of these guys here are constructed. Because in the first
instance this category is one that the kids administer. It’s a generic category,
available to whoever can bring it off, and that’s different than the category
‘hotrodders’ which is laying around in this conversation. Let’s consider that a
little bit. And in doing so, we’ll be moving more explicitly to a consideration
of some of the things involved in the phenomenon of ‘an automobile
discussion;’ what kind of topic that is, and what it is that can be accomplished
within it.

Now if we can take it that to some extent ‘hotrodders’ is a category that
is by and large employed by kids to characterize themselves, and whose use,
to some considerable extent, they enforce, and whose properties they enforce,
and obviously it’s, at least to some extent, a category that rebellious persons
can use, then at least one of the initial questions we might ask is: Why should
it be the case that at least some people who go about doing kinds of rebellion,
do it by formulating themselves as some particular type? That is, why do they
set up a type? Why don’t they try to make themselves observable as
‘individuals,’ so to speak? That might be an alternative to setting up a type.
(Of course ‘hotrodder’ is by no means the only teenager type that they set up
and enforce.)

That kind of a question - Why do they set up a type? - is not necessarily
a good question. What we want to see, in formulating the problem they have,
is what is it that they seem to see as the things they have to come to terms
with. And that seems to be in the first instance, that at least for certain
activities which are rather important to them - like driving - it isn’t the case
that there’s any big problem about what category will be used to classify them
by whomsoever happens to see them driving around. Now, that is, in
principle, a problem, and I've occasionally focused on more-or-less unique
solutions to it. 1 And for kids in cars, they apparently take it that there is - and
there’s probably good reason for that - a rather unique solution. And that is,
a kid in a car is ‘a kid in a car.’ If a kid is driving, he’s seen as a teenager
driving, and he’s seen via the category ‘teenager,’ compared to the variety of
things that he could be categorized as. His problem, then, initially, is that he
is in fact going to be typed; where, for one, the category ‘teenager’ is a category
owned by adults. Which is to say that what it is that is known about the
incumbents of that category is something that adults take it they know, and
know without respect to kids’ proposing that it’s so, or agreeing that it’s so.

The problem has some other aspects, which are perhaps more obviously
deep. And that is, it’s clear that one of the crucial things to the teenager is the

1 See Spring 1966, lecture 7, pp. 325-7.

Lecture 7

173

problem of independence, and one of the core facts about it is that in its given
form, adults set up and decide and enforce what it is that it takes for a kid to
be independent. And that poses for each kid a task that he has to solve on his
own, in confronting an institutionalized setup that’s going to handle him one
by one. Now the question is, is there some way of setting up solutions to the
problem of independence, which kids administrate? So that they decide when
you’re independent? Where adults will, in the course of time, simply be forced
to accept those characterizations, as they’ll be forced to use whatever
categories the kids invent to characterize themselves? I’ll consider the gains of
that kind of thing, to some extent. But it clearly would make a considerable
shift in the independence problem. For one, it would not be an issue where
each kid faces the adult world on his own. To set up a type, then, like
Hotrodder, or Surfer, or Beatnik, if it can be successfully done, is then to get
for the collection of persons a very large gain. That is, if the ‘they’ group,
whoever it be - call them ‘adults’ - come to use the type as well, but use it
under the extremely important constraint that what it takes to be a member,
and what it is that’s known about members, is something that the members
enforce.

Now a first thing that one wants to be able to do, is to so construct
appearances, and to so let out information, that members can take it that
when they’re seen by whomsoever, they will be seen as a member of the
category they want to be seen as a member of. That is, they’re not going to
be seen as ‘teenagers in cars;’ they’re going to be seen as ‘hotrodders’ - or
‘beatnicks,’ or ‘surfers,’ or whatever else. That, of course, poses a series of
tasks. What is it that it takes to get a category like ‘hotrodders’ across?

There are some things that turn out to be quite obvious solutions to that
problem. Everybody drives cars, and ‘teenagers driving cars’ involves nothing
more than looking into the car, whatever car it is, and seeing that there’s a
teenager there. Now the thing about hotrodders’ cars is that you don’t have
to look into the car. The car itself is so constructed that at any distance you
might choose, you see ‘a hotrod.’ And of course what’s crucial to that, in the
first instance, is that other people do not play around with the products they
get. They take the car and they drive it.

The modification characteristics have some rather usual features for such
kinds of categories, and that is, anybody can tell it’s a hotrod, but it is
members who can tell if it’s a good one or a bad one, what rank it has, etc.
And in that way of course, there’s strong protection against outsiders coming
to see how it is that whatever ranking goes on, goes on — except by, in fact,
becoming members of a sort. That thereby gives much freer play to the
members’ deciding and enforcing what counts as a really good hotrod. Now
there are, of course, rather classical reversal procedures, i.e., what looks like
the worst car to anybody else, is the best car. And those kinds of things are
very frequent in any such kind of operation. Presumably it’s just focussing on
that thing alone, that ‘turn the world upside down’ in that sense, that makes
almost all of, let’s say, the Beatnick characteristics. Hotrodders do not stand
in that sort of simple opposition; it’s not a pure Hegelian operation.

174

Part II

Now if it’s the case that what goes into deciding the rank of a hotrod is
something that kids decide, then one can at least begin to see how it is that
they’re able to set up, and what are the kinds of things they would use to set
up, machinery for social control over candidate members. That is, one is a
member by recognition of others who are members. And thereby, to
successfully get membership, you have to do what it is that they provide is the
way to become a member. You don’t ask your parents for permission, and
then treat that as your entry card. Nor is it something you can do without a
very considerable commitment. It is not something that you can do on an
occasion, in any given apparatus. That is, you can’t take your parents’ car out
for a weekend and go hotrodding. There are a lot of things you can do with
it, but that’s not one of them. And adults are, by and large, excluded from
the use of these things.

That has some rather important kinds of aspects to it. It permits one of the
crucial things the kids are concerned to get, and that is, an initial equalization,
such that the determination of one’s status as a member is not decided by as
many relevances given by the socio-economic structure as can be excluded. It
really is not any advantage to have, let’s say, the fastest stock car to start out
with - and in any case, it would not be ‘hotrodding’ to drive it. And it’s very
important, of course, that kids do not allow it to be ‘hotrodding’, and that the
way to not allow it is simply not to recognize the moves of someone who, for
example, pulls up to you in such a car. That kind of thing is very regularly
done for all sorts of other things. Years ago, when sports car drivers used to
wave at each other, there were all kinds of rules about what kind of car is a
sports car; like a Yolkswagon would not be waved at. Volkswagons could go
down the road waving at everybody, and get no response. Now what has to
be seen is that the response is not anything that is enforced in some way, like
if you wave at a Volkswagon somebody comes up to you afterwards, or
anything like that. The whole group has to be defended, each one by himself.
And in that way, Joe’s hypothetical little tale of the other fellow in a Pontiac
station wagon is a kind of mutual delusion - if there is such another fellow;
that is, they may be able to go through it together, but there is no proper
member who would accept either of them. What Joe is proposing, I’ll
consider eventually because for them it’s an extremely important kind of
claim he’s trying to lay out.

The character, then, of an attempt to set up a category like this is - let’s
say within Western tradition - a classical attempt at how it is one goes about
doing rebellion, at least the first feature of which is that one sets up a category
you administer yourself, which others come to use, and come to use in just the
unique fashion that they used whatever category they used on you before.
Where, initially they used another category on you, uniquely. They didn’t say,
“He’s middle class,’’ and “he’s this and that,’’ and “he’s such and such;” they
said “He’s Negro.” And the same holds for the teenager. Now of course, that
leaves those who don’t join, to be so called.

Lecture 8

Invitations ; Inexhaustable topics;
Category-bound activities

Given that we have gotten into a position where some aspects of how this
group is organized, and how that organization is demonstrated, how they use
it, etc., have been considered, we can return to the first part of this segment
and dig some more out of it, which we’re now in a position to consider, and
which can lead us further, also.

Joe : We were in an automobile discussion

Henry : discussing the psychological motives for

Mel : drag racing on the streets.

In our previous discussion of this segment we examined its collaborative
aspects. There’s much more there than that, and I’ll deal with some of those
things now.

We said earlier that that first line could have been a sentence if it was just
left as such. And that in not leaving it, Henry moved to use the structure of
the sentence to give some indication that these people were organized. That’s
grossly true, but there are other things about it. For one, if we consider the
first sentence as a sentence, then what we want to ask right off is, what is being
done by it? Among the things that could be done by it, one obvious thing is
the following. Someone has just come in to this group, they’ve exchanged
introductions, now there’s this first line. And it’s a very familiar first line, and
what it is, is, a possible invitation. A bunch of people are sitting around and
talking, somebody comes in, and they tell him what they’re talking about;
where they can formulate that in such a way as to provide that it’s something
that he could be talking about. It’s a topic for such as we, without any special
classification of who ‘you’ and ‘we’ are, just by virtue of the fact that we’re all
teenagers. And with regard to a consideration of the personnel, you’ll find in
the rest of the conversation that there are several other places where Joe makes
a move like that. For example, after they’ve worked Bob over, Joe introduces
a question that is an invitation: “What school did you go to?,’’ and there may
be others like that.

What I take it they do is shut that invitation off by proposing, in part, that
we’re talking about that thing which you know about, as any teenager would
- but we’re not necessarily talking about it in a way that you’re informed on.
But then there’s the addition to it, which is again kind of mundane and
ordinary. You can make up sentences like this. Here’s one - it’s not exactly

175

176

Part II

in point, though I think it’s pretty close. Suppose there’s a bunch of girls -
and we don’t need to specify the situation - and somebody comes in. One of
them says “We were in a discussion about sex.’’ Then we get “discussing the
physiological side effects of,” “using the pill,” as a third. The first and third
are something that somebody might well figure that they could, just knowing
that, come into. The second may, however, be something which says, this is
something special and unless you have the special information that you don’t
necessarily have, wait before you jump in. It’s the modification of the topic
that undercuts the status of the preceding part as ‘an invitation.’

Given the kinds of remarks we so far made - and more I will make - as
to the fact that the group is organized, we can get some pretty good idea
about, on the one hand, an invitation should be made right off, and on the
other, why it should be cut off. If the group is organized in a two-to-one setup
(or even one-one-one, as long as it’s not three-to-zero), then it would be quite
crucial to the members who this new candidate joins up with, and it could,
of course, be quite crucial to him who he joins up with. And if he simply
jumps into the conversation - where it’s how the conversation goes that
provides those affiliations - then some of the kinds of resources that would
permit the sort of decision some of them want him to make, would not be
available.

He doesn’t yet know that whatever he says is going to be treated as
something which constitutes an affiliation to one or another side. And if it’s
the case that even before this actual segment had got going they knew what
the structure was, and it was two-to-one against Joe, then Joe’s best chance is
to get this guy in, sight unseen. He might go with Joe. If so, things are
perhaps equal, and if he doesn’t, maybe there’s nothing much lost, in that
once the group structure becomes available it’s probably not unlikely that, if
he makes his decision on that basis, he’s going to go to the two others. And
that’s not unlikely, I’d say, not on just the sheer fact that any given person,
if there were a two-to-one situation, would go with the two, but one of the
obvious things about him is that he’s scared. He’s scared, and that’s why he’s
here in part, and also, he’s scared now. And under that circumstance it’s likely
that he’ll pick the safest way out; which he does eventually do. That is, the
first remark which can readily enough be seen as taking a position on his own,
is between 3:00 and 3:30, when Joe does one of his not infrequent coughs.
Bob comes on with “Sounds like an old man.” That’s about as dramatic a
remark as he can make in this. So in the collaborative sentence, then, it’s not
the case that it really wouldn’t matter who it is that did any of the parts. It
does matter, and the part that each of them chooses to use is not terribly
puzzling.

There are some things about this segment that are worthy of at least further
thought. But let me first make a remark about the segment that follows. The
most extraordinarily interesting thing about it is the way in which the
questions are informative. They say to him, “You must be crazy or you
wouldn’t be here.” It’s a terribly important beginning because it says ‘We are
that.’ Every question is a piece of information about the character of them;

Lecture 8

111

what it takes to be somebody who’s here - at least formulated in terms of
what an outsider might say about them.

And it’s very important that this question-answer series comes out right.
If it doesn’t come out right, it’s not simply, as an obvious consequence, that
he doesn’t belong there, but it may well equally be that they don’t. Because
if he’s there, however it was decided, he could belong there. And if he’s not
crazy, then, maybe this is not a place for such, and if it’s true about them, then
they don’t belong; where there’s a notion that the people here ought to have
problems they have in common. (I’m not giving any normative formulation
to ‘problems.’ They’re called ‘problems,’ and they would say they are called
‘problems’ without necessarily buying into that on their own part.) But if it’s
the case that they go through this procedure of asking a set of questions, and
they’re right, then they get a very gloomy finding, and that is that the society
is right, because these are a list of complaints that are made about them. If, in
fact, they locate whomsoever happens to come in here, then there’s some sense
to the fact that they know what they’re doing in putting people in such a place.
So the procedure is an extremely important one, however it may come out.

Let’s go back to this first segment. They’re in an ‘automobile discussion.’
Now here’s a fellow coming into this therapy situation, very likely his first. It
could be supposed that he has no idea what it is they would do in there. If
they wanted to put him on, they could name any topic they pleased, and he
probably would have to figure that that’s what they might be doing. When
‘automobile discussion’ comes out, it could be a very nice thing to hear. “Oh,
they talk about whatever it is kids talk about anywhere, and they’re not going
to be talking about things that are really godawful, or for which I just don’t
have any information so that I could partake.” Except it’s much worse than
that.

We get the modification which says hold on for a while, see what we’re
talking about and how we do it. And then we get a sample. And the sample
can be seen partly as a kind of ostensive definition of what ‘discussing the
psychological motives for drag racing’ means, and how a topic that you can
name in any scene that kids happen to be talking in, gets handled in this
scene. In which, we could say grossly, that what might appear to be as safe
a topic as you could pick, turns out to be altogether unsafe. And whatever
position you take is nonetheless usable to find out what’s bugging you. And
I suppose as strong a demonstration as needed to show that once you’re here
there’s no place to hide, is to pick as mundane a topic as you could name, and
then use that to carry on the intimacies of examination that they do.

When I use the term ‘safe’ for ‘automobile discussion,’ I’m not simply
saying that. That’s how they would quite consciously formulate the matter.
At one point in another session they’re for some reason or another exchanging
fantasies about something they’d like to have. It’s about maybe Bob’s second
session there, and he says he’d like to have a “’64 Triumph Bonneville,” and
he gets a lecture from Henry as to the fact that that’s not the sort of thing to
pick; that you don’t have to pick anything that sounds as ordinary as that, and
if you think that by picking something that sounds safe you get away with

178

Part II

anything, you won’t. (Which is not to say that as far as any of them might
well understand it, fantasy is about such things.)

I just want to extend the matter a little further, and simply suggest a kind
of problem about which I really don’t have very much to say, but which is,
I think, examinable. Consider things like the topic ‘automobile.’ One of the
things that an outsider, an adult, somebody who is not American, expectably
says is, well that could be a topic, I suppose, but it would probably be readily
exhaustable; how much is there to say about it? And you could imagine that
they could construct such a notion of cars as a topic by considering, for
example, that there are a finite set of parts to cars as facts, and that, as you
might exhaust the parts if you were assembling it or taking it apart, then you
could exhaust the topic. That is to say, you could have an image of eating into
a topic as you discussed it.

Now there are lots of topics for any sets of persons, but within cultures
there are topics that are intrinsically rich, in the sense that whatever it is that
members of that culture tend to talk about - that is, whatever themes they
talk about - they can talk about via that thing. For example, an automobile.
If one sat down to make a list of what things members of this society can
talk about when they’re talking about something, then one might find that
for kids, automobiles are like that. For example, ‘independence.’ You could
make a list of themes like that, and then see whether or not a conversation is
had, where that’s the theme that gets focussed about automobiles. If this is so,
then you have a rather decent way of saying that there is to some extent an
independent culture operating. And that the automobile has now become a
focus for kids in the way that whatever else is a focus for the adult culture -
if ‘adult culture’ is a thing to talk about.

Of course it might not be the case that some other culture has a single
focus - if cars are a ‘single focus.’ What I’m trying to say is that one of the
things one has to do in examining this phenomenon of automobiles as a topic,
is to formulate the character of that object within the set of things which can
be discussed, for whatever you’re going to formulate it that people are doing.
And then of course it would be quite irrelevant to see that adults can’t or
don’t or won’t say such things about cars. Cars are now an object similar to
parrots for a society in which a parrot is a god. And you wouldn’t say, “Well,
what can you say about a parrot?’’ or “What kind of interest could people
have in the difference between this parrot and that parrot, or this history of
parrots, or whatever else?” You would just not know what they’re talking
about.

You get, then, a way of seeing the kind of split that would take place when
an object is made into a sacred object - and a quite mundane object can be
made into a sacred object. Where, for others, it’s a purely secular thing: You
take it as you get it, and you drive it. And whatever happens to it, so what?
You can always get another. And also, of course, you get a way of
understanding that persons who deal with this as perhaps sacred have feelings
about the miserable way anybody else deals with it. But I don’t really know
that automobiles are a first-order object. It could be, because most of the

Lecture 8

179

conversations I listened to, I found so terribly boring that it’s only recently
that I’ve gotten up enough strength to sit and listen to them again, when they
go on for six hours at a shot, talking about carburators.

There’s a classic story about Malinowsky going into the field, wanting to
find out about geneology. He would talk to the natives about geneology, and
he found out that each native would talk for an hour before giving him any
geneology, in just bragging about them, and he’d have to go through a whole
big thing. Then at one point he remarks that he really became an
anthropologist when he could realize that geneologies were irrelevant, and it
was the talk, the bragging, etc., which was crucial. And you have to learn to
see that happening. I used to sit and listen through these automobile
discussions for the points when they would overtly talk about their troubles,
and I just don’t think that was at all right. It may be its partial distance from
somebody like myself that makes this a really good means for examining
what a topic like that is, and what you would be talking about when you say
it’s inexhaustably rich.

Now I want to move to considering the argument that Joe develops,
because one of the things we come to see, then, is what kind of category is that
category ‘poor little rich kid’ which is thrown at him when he develops his
position. But I can’t do that in the time we have left today, so what I’m going
to do is something that I could have either taken the whole course to do, or
I’ll have to do in ten minutes; it’s either that complicated or that simple. So
I might as well do it this way.

One of the clear problems that I can propose from this conversation is, we
could ask: How does it happen that discussions which become arguments end
up in name-calling? That’s a very standard observation, that it happens. It
happens here. Now, how does it happen? There are several things involved in
the formulation. We’re assuming that the discussion becomes an argument,
and we’re not asking how does the discussion become an argument. We’re
saying that once it becomes an argument it can end up in name-calling, and
that ends it.

The way we go about handling it involves some very general kinds of
observations which only become really interesting when we work them in
detail - which we won’t get to do.

One of the ways that a problematic occurrence is resolved, is by assigning
to the doer of it, some category about which it can be said that the activity
done, is ‘bound’ to that category, i.e., if you knew in the first place that he was
a such-and-such, it wouldn’t be any problem as to why he did the thing he
did. ‘They’ do such things. Now that fact has consequences in all directions.
One of them is, for example, if a problematic occurrence has happened, and
one knows a category that’s bound to it, i.e., where that category would be
said to be the person to do the thing, then you can construct a search
procedure for finding who, in fact, did it: Look to the set of people who are
so categorized. You could also, apparently, determine that any person who is
proposed to have done it, did it or didn’t. If he isn’t a member of that
category, then he wouldn’t have done it. And that has all sorts of uses, and

180

Part II

there can be lots of shifts in it. So, for example, the basic character of the
argument about Oswald’s murder of Kennedy turns on the issue of some
people saying, “Of course if he’s a Communist he would have done it,” and
others saying, “The last person who could have done it would be such a one
and therefore he couldn’t have done it.” To which, as I mentioned earlier, his
mother had the best solution of all, when she said that he’s a human so he
could have done it - which is to say that it’s not something that is to be found
problematic in the way everybody’s been posing it.

The fact that some activities are bound to some categories is used, then,
in a tremendous variety of ways, and if somebody knows an activity has
been done, and there is a category to which it is bound, they can damn well
propose that it’s been done by such a one who is a member of that
category.

What’s important, in part, is that it’s not the case that deviant activities are
especially problematic, but there are categories of persons who do deviant
activities and you’ve got a solution to a deviant activity if you’ve got a
member of a category which is known to do this. I’m not going to deal with
the problem of exceptions, although it’s dealable with and there are some very
nice things about it; I’m just not in a position right now to lay it out. It is the
case, though, that exceptions just don’t matter. That’s easiest to see by seeing
that the first Negro and the first ten Negroes you know, can be seen by you
to be exceptions to what you know about Negroes. It’s not the case that
exceptions involve any change in what you know about the category’s
members. For all the categories that have such kinds of characteristics as that
there are a bunch of activities bound to them, exceptions don’t matter. It’s
built in that there are exceptions, and they just don’t affect what you know.
You know that category does the following, and you know that there are
exceptions, and they do not involve you in modifying what you know. I talk
about that as: All these categories and the things that are known about them
are ‘protected against induction.’

The question is, what makes a problematic activity? And an utterance, of
course, can be a problematic activity. One of the simplest ways that an
utterance can be a problematic activity is, if a set of them has been made by
a set of persons and they’re not consistent. That is, there’s a position 1, a
position 2, etc. Now if, let’s say, you have two positions that are inconsistent
to each other, then it may be the case that you can make one position stand
by simply removing the other; like providing it’s problematic and explaining
it. The other then stands as the only one left.

Now then, one of the ways that one goes about making a problematic
position accounted for is, for instance, to assign some category to its sayer. The
category is, of course, important, because there can be some categories that
are, for example, the categories of persons who are entitled to say such a thing.
So that if you say that some scientific assertion is made by a scientist, that
would not explain it away, but will provide that it’s correct. But for a great
many, if you assign a certain sort of category to it, then you can provide that
the thing itself can be explained away. And that simple operation has been the

Lecture 8

181

basis for one quite sophisticated sociology. Mannheim proposes that any
sociological assertion can be given that formulation. That is, there are a set of
categories in the social structure, all of which are competing in that social
structure: Upper class, lower class, bourgoisie, worker, whatever you want.
Take any sociologist’s assertion, and say he says it from one or another of those
categories, i.e., assign a sociologist to a category, and you’ve made his
statement subjective.

Lecture 9

Character appears on cue ; Good
grounds for an action

I’m going to start considering : 1

Joe\ In that Bonneville of mine. I could take that thing out and if I’ve got
a tie and a sweater on and I look clean? ninety nine percent of the
time a guy could pull up to me in the same car, same color, same
year the whole bit, roll up his pipes and he’s in a dirty grubby tee
shirt, and the guy’ll pick the guy up in the dirty grubby tee shirt
before I - he’ll pick me up.

This is an extraordinarily complicated statement and we’ll attack it over a
while, I suppose. Let me just begin by noticing a relatively minor thing.
When we get this thing beginning with “In that Bonneville of mine’’ we have
what is a relatively rare, but standarized occasion, and that is that we can say
right off that ‘Bonneville’ is going to be the N of some N — *■ P sequence. It
looks for sure that that thing is not going to be dropped, and that we’ll then
get ‘it’ or something like ‘it’. In our discovery of the rule, I said normally we
do not start with the N, we start with the P. But in this sort of case, we can
pretty well be sure that we’re going to have the reverse rule apply. That, of
course, involves the fact that we get this kind of prepositional phrase which
would follow , e.g., “I could go out ... in that Bonneville of mine.’’

Now, generally throughout this statement we see what you might imagine
to be a kind of queer way that the whole thing gets assembled, i.e., clause by
clause. There’s no sentence in it anywhere - though one might consider from
“ninety nine percent of the time ...” on, to be an adequate sentence - but
otherwise you get a buildup (I’m not using that in the same sense I used
‘buildup’ as a technical term previously) in which, at the end of any given
clause it’s nonetheless the case that we still haven’t got a sentence, and we
clearly still haven’t got the end of an utterance. To what extent he uses clauses
(and there are much more elaborate things he does) to preserve the fact that
you can’t stop him at any point and have heard what he’s going to say, that
may be one of the things he’s engaged in doing here.

(If you’re interested, I can mark it for the way he said it: “In that
Bonneville of mine.” There’s a pause. “I could take that thing out” pause

1 The fragment is transcribed from the lecture and is not exactly the same as the
version in the handout.

182

Lecture 9

183

“and if I got a tie and sweater on and I took clean’’ pause “ninety nine percent
of the time a guy could pull up to me in the same car” pause “same color
same year the whole bit” pause "roll up his pipes” pause “and he’s in a dirty
grubby T-shirt” pause. And then the rest of it runs out to the end. Then he
starts up again, but gets cut off.)

One of the things one has to examine about that statement is, how is it that
it’s nonetheless a quite understandable utterance, i.e., despite its grammatical
oddness and despite (if you want to use the word ‘despite’) some other
features it has. For example, the phrase “ninety nine percent of the time”
occurs in juxtaposition with “a guy could . . . etc.,” except I don’t take it we
hear it as modifying “a guy could . . .” If you were to put it in sequence, it
would be: “ninety nine percent of the time . . . the guy’ll pick the guy up,”
and we nonetheless take that clause and put it where it belongs, when we
come to understand it. So there is, then, a considerable freedom in the
positioning of the parts of this statement, where that freedom apparently
doesn’t undercut seeing what it is he’s saying.

It’s also an enormously condensed statement. And that’s one of the things
I want to begin to focus on. Let’s take as a given, for the moment, the fact
that in “the guy’ll pick the guy up,” the first reference doesn’t refer back to
the prior “a guy could pull up to me . . . ,” though that’s a perfectly
expectable thing, but that we see perfectly clearly that it doesn’t, but that it’s
a new fellow, and that it’s a cop.

Now, if you remember back to the beginning of the course when I
considered “The baby cried. The mommy picked it up,” one of the things I
was remarking there was that when a character who has some proper grounds
for occurring and some proper thing to do, has its cue, then there’s no need
to account for how they happened to have come on the scene. “The baby
cried;” you don’t have to get a characterization of where the mommy was,
that she was anywhere around; the mommy does her job. 2 You can notice
again here that the cop is not introduced, as, for example, is this other guy
who is driving - he pulls up, gets described, then we have this bit. What
occurs is good grounds for the cop to do what he ought to do; he’s on the
scene and he does it. So he’s introduced via the action he does, where the
grounds for that action are laid out, though how he happens to be there need
not be indicated.

And the question of what character doing what action that kind of thing
can hold for, is quite non-trivial. To provide for this introduction of the cop
(and, for example, that introduction of the mommy) being quite unprob-
lematic, we would have to work out analyses which would involve building
the characteristics of a class of actor - in this case, ‘cop’ - as a staff deputed
to do a set of jobs, and describing how they go about doing those jobs - for
example, by just driving the highways - and then considering the information
about such a class of actors that any member would have, such that it would

2 This particular consideration was not captured in Sacks’ notes for lectures 1 and
2. For a similar consideration, see Spring 1966, lecture 2, p. 254.

184

Part II

in no case be problematic that on a single occasion of the cue occurring, the
character occurs. (This is not to say that the sheer fact that there’s a norm that
says speeders ought to be arrested is sufficient to provide that the fact that he’s
speeding brings the cop on the scene.) For organized crime it would be quite
different. A bank robber would be much puzzled how come when he arrived,
the cops were there waiting for him, and he would ask, “How did you know?”

One can presumably consider this sort of phenomenon and perhaps get an
idea of what the sense was for, let’s say the Greeks, who introduced in the
tragedy, the appearance of a god at some point. We might come to see it, not
as what we give it the sense of, i.e., as absolutely unwarranted or unexplicable,
or just happening, but as something that was somehow institutionally
provided for. We would want, of course, also to see, then, what sorts of
proper occurrences need an account of how a character who is doing what’s
expected or proper, comes to be there. We could perhaps dig out the different
classes by just examining things like this statement of Joe’s, where it’s
absolutely unproblematic that he doesn’t introduce the cop as sitting behind
a billboard, or whatever else. The cop is introduced via his action.

And in this case at hand, it’s a rather more powerful kind of thing than
even that. In “The baby cried. The mommy picked it up,” we had the
specification of what was in fact good grounds for the mommy to do her
action; “the baby cried.” In this case, that, too, is just left out. It is not the
case that rolling up his pipes is what’s good grounds for the cop to arrive. It’s
that rolling up his pipes is the first move in a sequence which the occurrence
of that technical term provides as going to happen. The driver pulls up,
signals the other, they have a kind of drag race, and the cop catches them in
the middle of it. That part is dropped out altogether here; filled in by the
hearer.

Now it’s very nice to have the hearer fill it in because one of things that that
will guarantee is - and this is an important part of this fellow Joe’s claim -
if you’re going to fill it in, you’re going to fill it in with what would be
legitimate grounds for that arrest to take place, and not with some quirk or
some erroneous arrest. Then the question of what can be left out of an account
- especially where it’s the thing that’s left out that provides for the crucial
action to happen - is another rather crucial thing to consider. For one, that’s
the way that the condensation of an account can take place. The work of
digging out what’s being said is, then, the task of the hearer, and one can
leave out what one knows the hearer has to fill in in a specific way. That is,
the hearer is not going to say “Rolling up his pipes is nothing to get arrested
for,” but there is some sequence, of which that is the first move that occurs,
and when the actionable move occurs, then the arrest takes place.

In fact, it’s not simply the case that a hearer can fill in such a blank. They
can do that, but they can do a lot more. Here is a piece of conversation which
has the following kind of setup. A fellow is calling a counselling agency in
regard to the trouble he’s having with his wife. It’s quite early in the
conversation; a first conversation between two people who don’t know each
other. ‘A’ is the staff member, ‘B’ is the caller.

Lecture 9

185

A: Yeah, then what happened?

B: Okay, in the meantime she says, “Don’t ask the child nothing.” Well,

she stepped between me and the child, and I got up to walk out the
door. When she stepped between me and the child, I went to move
her out of the way. And then about that time her sister had called the
police. I don’t know how she - what she —

A: Didn’t you smack her one?

B: No.

A: You’re not telling me the story, Mr B.

B: Well, you see, when you say smack you mean hit.

A: Yeah, you shoved her. Is that it?

B: Yeah, I shoved her.

Again, it’s not like they’re old buddies. The secondary characters here are just
named categories, i.e., ‘her sister.’ He doesn’t know her sister, doesn’t know
if she’s short-tempered or what, doesn’t know how B and his wife get along,
or what B is wont to do if his wife does something. Nonetheless it’s not simply
the case that he has some idea of what it is that would provide for the sister
to call for the police, but when he’s told it’s not so, he takes it that there’s a
lie going on, and he’s in a position to insist and not simply to guess at what
it is that transpired. In that sense, then, knowing what are good grounds for
the police to come is something that permits him not only to fill in a blank,
but also to deny a statement which proposes something else, with no need to
specify anything more than ‘me and my wife,’ ‘her sister,’ ‘the police.’ What
is in operation here is the pure categories, and their organization, and what’s
known, given their use. The guy who calls is in no position, apparently, to say
“How do you know?,” “I’m telling the story,” or whatever else. It’s not
simply then that the guesser feels quite assured in proposing what it is that
happened, but the other guy is just as well in a position to see that the guesser
had good rights to do that. I take it, then, that one of the things we would
not be saying is, it’s a matter of style whether one leaves out a sentence that
might be put in.

How that can happen is of tremendous importance, and it tells us a great
deal about what it is that, among a set of persons who don’t know each other,
nonetheless can be taken as so about how the world operates, in its
particulars. That is, on a single occasion. It’s a different kind of business than,
“If you go hotrodding you’re bound to get caught.” It’s on this occasion the
cops arrive. How come?

Now, with such a guide, you could begin to see such issues as, what the
task would be of a foreigner to understand that thing. Suppose you were to
translate this. The issue would not be anything like, do you come up with a
word-for-word translation? What you have is do is build in characterizations
of the way a set of institutions operate, the way sets of persons know what
anybody else knows, and the like. Just to provide for a quite ordinary segment
of conversation.

The big question is how we come up with clean ways of formulating what

186

Part II

the kinds of condensation are. You could ask it in a fashionable way: What
would it take to have, as the staff member in that last conversation, a
computer? What would it have to know, to be able to engage in a little
conversation like that? If you could build computers that could gossip, then
you would have a formal characterization of what a culture was. And the
character of what these condensations are, is not to be supposed as satisfying
any unexamined notions of what, let’s say, we take to be prototypical kinds
of condensed talk, like talk in mathematics. That’s in part the problem with
Reichenbach’s seventh chapter, where he simply will suppose that when
persons use terms that have a known logical sense - like ‘and’ - then that’s
the sense they have.

Returning to “the guy’ll pick the guy up.” Earlier I mentioned this notion
that some activities are bound to some category. And I take it that when we
hear “the guy’ll pick the guy up” we use ‘pick up’ as an activity that ‘cops
do, to provide that ‘the guy’ is ‘a cop.’ And we don’t find ourselves in the
position of being confounded by the antecedent rule, which would say ‘the
guy’ has to be the fellow introduced as ‘a guy,’ since ‘a’ followed by ‘the’ is
a perfect way of indicating that the same person is being referred to.

There’s a kind of elegance to the use of ‘the guy’ there. It is in the first
instance a tremendously awkward kind of way to say the thing, “the guy’ll
pick the guy up.” To have it that ‘the guy’ here is not the person who is
introduced as ‘a guy’ earlier, would seem to be generative of further trouble.
But once he’s got this ambiguity going, he has an opportunity to reintroduce
one of the crucial assertions of the whole segment, and that is, “in the dirty
grubby T-shirt.” So that grammatical awkwardness provides the occasion for
emphasizing what he’s trying to emphasize, where, were he to do it in the
grammatically proper way, such an occasion would not be available as it is
here.

And we can find that ‘dirty grubby T-shirt’ is something that is
emphasized, independent of its being repeated, in the following ways. First of
all, the whole statement is hypothetical. Every clause in it is hypothetical. In
the whole statement, there is one use of the most standardized way of making
a hypothetical statement, that is, the ‘if ’ in ‘. . . and if I got a tie and a sweater
on ... ” What that single usage seems to involve is, given that there’s a
whole set of things that are introduced as precedents to someone’s getting
picked up, how is it that he’s going to focus on what he takes to be the crucial
differentiating thing for that happening? It seems that he marks that crucial
contrast with ‘if.’ Where, then, we can see as parallel construction, “and if I
got” and the subsequent “and he’s in;” that is, we get not simply the
difference between my being clean and his being dirty, but the parallel
construction (‘and if,’ and then again ‘and if,’ with the ‘if’ dropped out - and
you don’t have to have the ‘if’ in such a parallel construction) sets those
things up as what differentiates.

Furthermore, ‘dirty grubby’ is an emphatic construction in itself, in a
variety of ways. First of all, it’s a double adjective; the same thing being said
twice, which is kind of characteristic of kids, and characteristic of this fellow,

Lecture 9

187

also. Then, it’s not just a doubled adjective, but a doubled one that is
soundwise emphatic. ‘Dirty grubby,’ ‘teeny weeny,’ things like that.

There are other things involved in this, as well. ‘Grubby’ happens to be the
way that one of the members of this group gets referred to. A way they get
at Henry is by characterizing him that way, and apparently he does, indeed,
tend to be grubby. On page 5 of the eight-minute segment, he says, just
before 6:30, “I’m not grubby’’ and that brings a laugh. Now, I’ve been
leaving that sort of thing out; that is, the way that in any statement which is
not at all specifically referring to somebody in this scene, they are nonetheless
doing that. That would have to involve providing much more transcript than
the eight-minute segment.

Lecture 10

Clausal construction; Hotrodding
as a test

I’m going to continue considering this thing:

Joe: In that Bonneville of mine. I could take that thing out and if I’ve got

a tie and a a sweater on and I look clean? ninety nine percent of the
time a guy could pull up to me in the same car, same color, same
year the whole bit, roll up his pipes and he’s in a dirty grubby tee
shirt, and the guy’ll pick the guy up in the dirty grubby tee shirt
before I - he’ll pick me up.

I’m not in a very good position to lay out how, in its details, this utterance
gets constructed of a bunch of clauses and phrases rather than a set of
sentences, or what each of the variations on what we might take to be the
normal form for such a set of assertions is doing - if it’s doing something. I
pointed out some things earlier, and I can point out a couple more.

One of the rather obvious facts is that, given this way of constructing his
utterance, the clauses are adding up and it’s not the case that at any given
point - even at the point where we have that question mark (“. . . and I look
clean?”) - that he’s said what he has to say. And that’s observable by the
hearer. The hearer doesn’t have to guess that he has more; it’s observable that
whatever it is he’s done, he’s not yet finished with it. There are some quite
clear ways that the fact that the utterance is still open at any given point, is
done by him. For one, whatever much we might have to say to give an
adequate characterization of the semantics of ‘if,’ it does seem clear that there
can be a multiple set of things being done in a statement that contain, at an
early point in it, an ‘if’ clause. It clearly is a marked first occurrence of some
sequence in which there will be a second part, the ‘then’ clause. There may be
an overt ‘then’ clause, or a clause that turns out to be a ‘then’ clause. Where,
then, as long as the ‘then’ clause hasn’t occurred, the utterance remains open,
and one can wait to examine any given further clause to see whether that is
the ‘then’ clause, and can close the thing off.

Or, for example, we could ask what would be the difference between
saying ‘‘I could take that Bonneville of mine out” as compared to what he
does, ‘‘In that Bonneville of mine. I could take that thing out,” where it’s
reasonable to say that neither would be heard as a complete utterance. Now,
given that we can see, when ‘‘In that Bonneville of mine” occurs there’s going
to be a second reference to the car, e.g., ‘it,’ or as in this case, ‘that thing,’ then

188

Lecture 10

189

perhaps what he is doing with this construction is having that second part
provide a contrasted reference to the earlier mentioned car. That is, “If you
take a big fancy car out’ ’ is what he’s introduced in his last statement. ‘That
thing’ can then be, perhaps, not only a reference to the Bonneville, but a
contrasted reference to the other - where a ‘big fancy car’ is a hotrod. That is
something that could be done with a clause like that, as part of a pair, ‘this
and that,’ ‘that thing’ ascomparedto ‘theother.’ and it would make part of his
argument, which is, “If you go out hotrodding with a hotrod you’re bound to
get caught” and in contrast, “If you take the Bonneville out, you won’t.” 1

There are, further, a set of ways that this thing is constructed which are not
exactly consistent with what is proposed to be the way that certain connectors,
most particularly ‘and,’ get properly used. A consideration of those can be
found in Fries’ book The Structure of English. His grammar is terribly
important because I guess it’s the only grammar of English that was
constructed by reference to an attempt to handle actual conversation itself.
Fries sat down with a bunch of telephone conversations and built his grammar
out of that. Quite unique in that way.

One of the things he proposes about ‘and’ is that the term on either side
are from the same form-class. That means, roughly, adjective ‘and’ adjective,
verb ‘and’ verb, etc. Seymour Chapman then extends Fries’ form-class use of
‘and’ to handle similary constructed clauses. That's in a paper called
something like ‘English sentence connectors,’ in a book edited by Marquand,
entitled Studies in Language and Literature. Now, what we have in this
statement of Joe’s, are clausal uses, except that they’re kind of queer. Take a
sequence like this: ‘. . . roll up his pipes and he’s in a dirty grubby T-shirt and
the guy’ll pick the guy up . . .” The consecutive clauses do not seem to be
similar, but they are separated, in a way. That is, there those pauses: “roll up
his pipes” pause “and he’s in a dirty grubby T-shirt” pause “and the guy’ll
pick the guy up . . . etc.” We could treat those pauses as putting “and he’s
in a dirty grubby T-shirt” in parentheses, and then we can notice that “roll
up his pipes” and “the guy’ll pick the guy up” are of the same class, and
‘and’ connects them, i.e., “roll up his pipes . . . and the guy’ll pick the guy
up.” The ‘and’ preceding “he’s in a dirty grubby T-shirt” is connected to
“and if I got a tie and a sweater on and I look clean,” and there it does the
same sort of work. So what we’ve got is non-consecutive uses of those ‘and’s,
which nonetheless do the same work that they would be doing if they were
consecutive - and how he comes about to break these things into this
non-consecutive fashion, I can’t really say. But what we’ve got is a series of
parallel clauses: “If I got a tie on” and “he’s in a dirty grubby T-shirt,” “I
could take that thing out” and “he could pull up to me,” etc. We’ve seen
earlier that this guy uses these parallel constructions very much, e.g., “you’re

1 Some of the class disagree with the proposal that the distinction is between
‘hotrod’ (‘big fancy car’) and ‘Bonneville.’ They take it that, as one of them proposes,
although the “Bonneville really isn’t a ‘big fancy car’ (i.e., a ‘hotrod’), this kid is
trying to pass it off as one.” To that, Sacks replies that if they had access to more of
the transcript they could see that the distinction holds.

190

Part II

bound to get caught” and “you’re bound to get shafted,” and a whole series
of others like that. What they are, I don’t have any idea.

In Joe’s first statement, he said “If you take out a hotrod and you go
hotrodding you’re bound to get caught.” Now what he does is to substitute
for the hotrod this Bonneville station wagon, and he here reposes the
circumstances of getting caught, where both persons are using “the same car,”
and what differentiates them, and what turns out to be relevant to getting
caught, is how they’re dressed. So we have a sequence of moves: It’s not
simply, now, that they’re not driving the hotrod, but even driving the
Bonneville there is a chance that they’ll get caught, and the way that gets
reduced is by what he wears.

What I want to do now is try to lay out, to some extent, what some of
the conflicts are between the guys in this discussion, and what things they
are proposing. I partially considered what the other guys are proposing a
couple of days ago. What he’s arguing, as I see it, is that your basic aim
ought to be to go speeding while minimizing your chance of getting
caught. And to minimize your chance of getting caught, there are a set of
constraints which are put on how you go about doing it. If you don’t use
them, you can’t win; so he proposes. From the hotrodder’s point of view -
if you can take the others to be espousing that point of view - such a position
misconceives what they’re up to. What they want is not a guarantee that they
won’t get caught. If they had that, they would have lost right then and there.
They want a situation in which they can expect to lose, because what that
permits is, if they happen to win, the formulation of explanations for having
won.

That is to say, if it’s the case that not many people get picked up in a
Pontiac station wagon, then if you don’t get picked up in a Pontiac station
wagon, there’s no problem which needs an explanation. If, alternatively, most
people who go hotrodding get picked up, then on any given occasion when
you don’t get picked up, there’s a problem which can be explained, and which
can legitimately be posed as, “How did that happen?” At that point, and into
that problem, can be fitted ‘skill,’ as an explanation.

It’s in that regard that why it is that one chooses an activity like
hotrodding, that involves violating some legal norm which is an important
legal norm, can be fairly clearly seen. What you need in order to get the
possibility of the proving of skill, is a situation where, if some given activity
has been done, ‘A’, then either ‘B’ or ‘not-B’ are both observable. That is to
say, both ‘getting caught’ and ‘not getting caught’ are observable things, so
that you can say that ‘not getting caught’ is something that happened, or that
‘getting caught’ is observably absent.

And of course an enforced legal rule is a rather usable object for this test of
skill. So it’s not at all the case that if they could be guaranteed that they would
not get caught, that’s something they would want. It’s a misunderstanding to
suppose that drag racers are simply playing each against the other. They’re
play against the cops just as much. And they need the cops, because they need
the cops’ catching or not catching of them to formulate that they have done

Lecture 10

191

the thing correctly; to even pose the fact that something has happened which
they could have done correctly. In order to have a situation where you need
an explanation, so as to be able to put in the explanation you’d like to put in,
then what you need is some situation where these B or non-B alternatives are
both equally observable things. If not many people get picked up, then if you
don’t get picked up, it’s just that not many people get picked up.

Now the confrontation with the police is not simply accounted for in this
way, though that’s a good part of it. There are some other things, and if,
when I’m talking about them, I use a kind of glorified language, let me just
say that it’s the language the kids use. These kids, and other kids who are
routinely engaged in observably deviant activities, regularly formulate the
problems in the same way. They start out with the fact that, in just cruising
the highways, they are observably deviants. That is to say, they are seen as
‘hotrodders in their cars,’ and they are seen as persons who either have just
finished, or are just about to, or in any case are routinely engaged in, speeding.
The rules of the game are that anybody who speeds can be arrested, but
they’re in a situation where, further, they know that the cops can, if they care
to, arrest them at any time. That’s not simply because the cops can say that
they were speeding, but because their cars, as with anybody else’s car pretty
much, if stopped and examined by reference to the criminal code, will be
found to have one or another technical violations. As they put it, “Last
weekend they were arresting people on Hollywood Boulevard for dirty
windshields.” And they know that anybody other than those who are
observably deviant will not get arrested, unless there is such an occasion as
speeding. What they can then examine is whether the police will conform to
the rules that the police ought to conform to, i.e., treat them as they treat
anbody else, or whether they’ll succumb to the temptation to violate the rules
of the game and make an easy arrest.

And in their observable deviance, then, they have a means of learning
whether or not people like the police are worthy of respect - whether they
conform. And if they don’t, then of course they can say, “They don’t
conform, why should we conform?,” etc., etc. In any event, you’ll find that
they’re terribly carefully oriented to the fact that these others do or do not play
according to the rules of the game. And the best way they have of testing that
is, of course, to put themselves in the situation where they give a deep
temptation to the others, to violate. Then if they don’t, you can say they’re
respectable. They get a view, then, of organizations like the police, which
others don’t have available. And it’s a very persistent topic as to whether, and
how, the police do or do not behave as they ought to, or whether they just
pick you up because they know they can pick you up - since, for one, if
anybody sees them pick you up, the others all have an explanation: You must
have been hotrodding. (Negro gang kids have exactly this same orientation.
They engage in long complex discussions, which are morality discussions of a
standard sort, as to how the police behave, and who is worthy of respect.) So
these kids have a way of checking out the extent to which that moral order
that they’re supposed to conform to is followed by those who are its enforcers.

192

Part 11

They’ve got a way to test it, they test it with perfectly good will, if they get
arrested for speeding, that’s the way it goes.

There’s really, then, a tremendous ethical conflict going on between that
position and what Joe is proposing - which is something like an adult
speeder’s view; that is, if you look perfectly square then you’re in a position
to get away with violations that others would never get away with. And he’s
proposing, then, to use that cover of his clean appearance, to generate
violations while being protected from being caught. For him, then, if he were
to do it, he would be gaining lessons in cynicism. It is also the case that what
he’s proposing is not available to any given kid; that is, he’s proposing to use
a set of resources that any given kid would not necessarily have. And their
concern is to try to set things up so that any given kid would have a chance
to participate in just the way that any other would, and, of course, to get a
view of how that legal order is operating.

The extent to which either of these alternative strategies get used, and what
consequences they have, are matters that I am in no position at all to say. I
have little bits of data like this, and not anything much larger than that. But
in very different materials one finds the same preoccupation. So it may be that
the conflict that’s set up here is not really terribly atypical - if that’s at all
relevant.

Lecture 11

Espousing a rule ; Exemplary

occurrences

Considering that statement, “If you go out hotrodding you’re bound to get
caught and you’re bound to get shafted,” there are some things about it that
we ought to note. First of all, the maxim that quite obviously follows from
it - that you ought not to do the thing - is itself not asserted. So that what
we get is the parts of a syllogism device as the basis for finding that maxim
as appropriate. The character of that maxim - that you ought not to do it -
is something that we could consider by reference to the fact that it isn’t
asserted, as well as by reference to some other features of the way it’s used.

One thing is, that with a maxim like that one, even supposing that
someone proposes it overtly, there are some alternatives as to what’s to be
made of it. For example, it could be treated as a more or less purely affiliative
statement, i.e., something prefaced by ‘I believe, and I am a member of those
who hold the belief that this is or is not a proper thing to do.’ Where, then,
an answer might be an assertion that those who hold that belief are members
of some group that we don’t affiliate to.

It is to be noted that in the variety of maxims that are implied by Joe’s two
positional statements, they are not explicitly proposed by reference to the fact
that he thinks they ought to be followed because they are, e.g., good rules, or
whatever else. So the advice he can be offering is the type which both the
offerer of, and the one who accepts, can offer or accept without any
commitment to those persons who have built them, those persons who
enforce them, etc. They’re only ‘good advice’ to anyone who lives in a world
which is at least partially governed by the fact that if you do ‘A,’ ‘B’ is liable
to happen. That is to say, Joe isn’t saying that you ought not to do it because
it’s wrong - where, if he said that, then they could simply claim that he was
a traitor among them, or perfectly clearly an adult in child’s guise, or
something like that. While the technique he’s using does not foreclose such a
possibility, it is very generally used. That is, the claim that some rule which
one does not claim is correct or good or whatever, ought to be followed by
reference to whatever it is are taken to be the interests of the persons being so
advised. Presumably it’s a major technique for social workers dealing with
street gangs and that sort of personnel.

In fact, the character of the warning or threat which is asserted via such a
technique, is the basis for the kind of phenomenon most technically explored
in the works of Thomas Schelling. These are a series of war-game type things,
one very classic technique, which has a weak form and then a much stronger

193

194

Part II

form, goes like this: You can listen on the radio to the announcements of
government leaders, and they will often make statements of the following
form: “If the US keeps on its present course,” say the Russians, “then it’s very
likely that war is inevitable.” Where they’re making what looks like a mere
prediction, the follow-up of which they have nothing much to do with.
They’re not proposing you ought not to do it, or that they think what you’re
doing is bad, or that they will have anything to do with those consequences.
It’s a natural fact of life. And for that kind of thing to get strengthened, the
classic military procedure is, I suppose, the kind of thing that was in the
movie, Dr. Strangelove , but can be found in places like Schelling, and that is,
to set matters up so that the consequences you propose naturally to happen,
will happen as an inevitable consequence of some event. So that if, for
example, a bomb explodes in a certain place, it will automatically and beyond
the control of anybody involved, set off a reprisal which has been electronically
geared to this possibility. Governments tend to do that sort of thing, and the
US, for example, uses the phenomenon of public opinion in that fashion, when
they propose that it’s necessary, given public opinion, that this or that military
strategy be enforced. It does not matter, apparently, whether the consequences
function electronically, or whether it is simply a matter of the fact that certain
events will have their emotional consequences for a public - allegedly or in
fact. And it’s not a matter, either, of whether you - whoever it is who’s
proposing this thing - are not the one who mobilizes that, and in fact can
ignore it.

So statements of this form seem to set up problems, the solution of which
consists of making them more and more descriptive in fact. That is, because
these are problems, if you can satisfy their conditions, then you’ve got a
powerful maneuver; so you arrange things so as to satisfy their conditions. If
you could have a statement to the effect that if the Soviets move their troops
in the following way, then, without respect to what you can do, the following
will happen, and they can see that it’s so, then of course it pays to arrange
things so that that maxim now holds.

And relatedly, it is important that if one is attempting to get a set of persons
to modify the behavior of some others who would stand in opposition - or who
certainly stand in no commitment - to the set of rules you hold, then you want
to be able to permit those who are going to espouse those rules to come on as
perfectly well understanding the circumstances of those they’re talking to,
perhaps even affiliating with them, but in any event, not at all committed to
the correctness or the moral rightness of the positions they’re espousing.

In, I believe it was the 6th century, the Catholic missionaries who went to
Catholicize England, wrote to the Pope saying, “These people already have a
religion, and it has all sorts of features to it; for example, they have places
where they worship, etc. What do we do about it, do we destroy them? After
all, they are heathenish places, and how in the world can they be accepted or
acknowledged by us.” To which the answer came back: “Leave them. In fact,
encourage them. What you want to do is fit our terminology and our ways to
whatever given ways these people have. We ll be around a lot longer than that

Lecture 1 1

195

one will, and over time they will have forgotten what the other religion was,
and we will nevertheless have maintained whatever power it has over them.”

What the constraints are on human nature, is something that we want to
discover via an investigation of how it is that they in fact do whatever they do;
where what we’re trying to do is to see what is it and how is it that some ways
they have of doing something works, and what its import might be. Now
here we start out with a routine argument of a fairly trivial sort: If you do X,
Y is bound to happen. There are reasons why it seems to be a quite powerful
one. We can get lots of examples of it. We can find that it’s used by persons
who hold to rules without regard to whether they’re efficacious or not, and
they offer the efficacious formulation to those who don’t hold the rule at all.
For one, that permits the ones it’s been offered to, to accept it and to
proselytize it without themselves changing their affiliation. If each person who
accepted the rule had to espouse it for its moral correctness, then each time
someone accepted the rule, he might now be expelled from the group he was
in, in the first place. Suppose you convinced a kid that he ought not to hotrod
because it’s bad. Each kid who was so convinced, might no longer have any
access to the hotrodders, who would treat him as a turncoat. What are the
circumstances under which somebody espouses a view that’s not held by a set
of members, and can still talk to them? Just consider the question that’s in the
air about these guys coming back from North Vietnam now. What are the
circumstances under which they could possibly be in a position to say the war
ought to be ended?

Now, one of the other things I want to examine for this is, it is perfectly
well possible that no matter how you happen to espouse a rule - that is, for
its moral correctness or its efficaciousness - you’re liable to face the
circumstances of being called a member of the group that espouses that thing,
and when you’re called that, the rule is thereby treated as irrelevant. The way
that happens has to do with the kind of things that, in their queer way,
philosophers have been much concerned with, at least early in the century.
And that is, as they put it, if you assert some moral rule, are you doing
anything more than asserting your affiliation?

Pretty much all that is interesting about that is, that it is indeed the case
that the way members account for and, often, explain away some asserted rule
is by aserting that category for which it is the case that its members hold that
rule. They can do that effectively when they propose to be members of
something that stands in alternation to it. And it is along such lines that how
it happens that arguments end up in name-calling would be laid out. Given
a position, a category is assigned in terms of which that position would have
been generated - where there are alternative categories. A situation is opened,
then, in which those members who have not declared themselves are now
warned that the positions they take will involve them in affiliating to one or
another of the categories that are now present. For each claim that Joe makes,
then each cateogry that’s offered as an account for those claims is being
proposed to Bob as, do you want to be called one of those? If so, take a stand
with this guy. If not, don’t say anything, or oppose him.

196

Part II

There are two different ways Joe goes about making his position. One is
that statement that has as its obvious inference the maxim ‘don’t go
hotrodding.’ The other is through this hypothetical singular occurrence,
which it seems he’s proposing to be fully generalized. ‘If I . . . etc.,’ where the
‘I’ is now merely a person in this thing, where what he says holds for
whomsoever. And that generalizability is of course attacked in the return to
it which says it’s not generalizable i.e., ‘Not many people . . . etc.’

We could call that second statement of Joe’s either an exemplary
occurrence or a hypothetical exemplary occurrence, and wonder whether it
matters that it’s hypothetical or not. And indeed, it needn’t even be clear
whether it is hypothetical or not, because he can use the ‘if form for an actual
occurrence, simply to show he’s going to make an argument from it; where
you use the ‘if form in order to make it stronger. That is, the sheer fact it
occurred may have no consequences. If you just say, “On a given day the
following happened to me,” the hearer can say, “So what?” If, however, you
say, “If I do the following ...” then you’re making of it a general argument.

Now the phenomenon of ‘exemplary occurrence’ is fundamental for the
following kinds of reasons. It seems to be the case that a large amount of the
knowledge that Members hold about how the society works are, as I earlier
put it, ‘protected against induction.’ And what that means is that it isn’t
automatically modified if events occur which it doesn’t characterize. The
question is, does that mean that there’s no way that it can get modified, and
if not, how does it get modified, where the question of how it gets modified
can be quite important.

There could be all kinds of models of how it might get modified. One
might be, you store up a bunch of exceptions, saving each one, remembering
them, and when you get a whole bunch of them, you’re forced now to say that
what you supposed were so is not so. For example, the first Negro you meet
turns out not to be what you know about Negroes. You store that up as
information. When you’ve met ten, and they all turn out to be not what
you’ve been told, you say, “Well I guess that information is incorrect.”

Now, things are not done in that fashion. Let me give an account of a
particular situation, so as to show what the consequences are, of their not
being done in that fashion. One time I was engaged in trying to see how it is
that coroners make decisions that somebody did or did not kill themselves.
Those things are handled case by case, one at a time. A case comes in - let’s
suppose it involves a body being found with a pistol wound in some particular
area of the head. The area matters, because the coroner can determine that if
the wound is in a certain area it could be an accident e.g., the person could
have been doing such a thing as checking whether there were bullets in the
gun, etc., whereas if the wound is in other places, then the person must have
had the intention of killing themselves in that way; that is, they wouldn’t have
otherwise been holding the gun in such a position as to lead to a wound in that
area of the head. They go through the cases, and each one may get decided
via one of these rules.

If you said to the guy who was making those decisions, “It’s a probabalistic

Lecture 11

197

argument; probably in five cases out of a thousand, it might well be that
somebody accidentally had the gun in that position,” he’d agree. And now
there could be a variety of ways of considering that five-cases-in-a-thousand
phenomenon. One would be to take a thousand cases of this sort, save them
up, then make your decision on all of them together, the five queerest ones
will be called accidents, the other 995, suicide. That’s not what’s done.
What’s done is, each case comes up. For each case there’s only a five-in-a-
thousand chance that it could be not-suicide, so each one is decided to be
suicide. Furthermore, when a problematic case comes up, what’s said about
it is, “There’s a very small chance that it could be an accident; it’s not very
likely, and furthermore we’ve always found that they are suicides.” And the
fact that, that they’ve ‘always found’ consists of their procedure for deciding
that, does not seem to make any given determination problematic. That
they’ve ‘always found’ is used to, in fact, urge that the next case is like those
that we’ve found. So you don’t get the adding up of the troublesome ones,
about which can be said, “Well, there are always some that are sort of
troublesome but they have always turned out to be suicide.” I take that
because it’s a rather stark example, but consider how people deal with
excuses, and a variety of other such phenomena; they do it in the same way.

Now then, the problem is how do these shifts take place in that knowledge
that gets used on each case, and for which the fact of exceptions occurring
turns out not to be relevant in each case; even though they will be fully
recognized as sometimes occurring. To extend the point a bit, consider for
example, those things which I’m sure have happened to you, where you know
that such things happen, you’ve heard that they happen, people tell you that
they happen - except that when they happen to you, you also know that you
could never tell anybody that they, in fact, happened to you. That is, they
happen. But they happen to nobody, on no given occasion. And you realize
you can’t tell anybody about it, except, maybe, to say that it happened to
somebody else.

Now the phenomenon of an exemplary occurrence is apparently a kind of
way that a large shift can take place. And in the case of suicides, just to
continue that along, things could happen as follows: Until now, deaths at
railroad crossings, when a car gets hit by a train, are presumed to be accidents.
Then somebody says, “We’ve had a whole bunch of these recently, they’ve all
been decided as usual, but this one looks interesting.” They examine that one,
and it turns out that it’s extremely unlikely that it could have been an
accident; the train could perfectly well be seen, and, further, the road was so
situated that the car could not have been seen from the train. Suppose also
that they do what has been done by people like those in the Harvard School
of Public Health; say, take the car apart afterwards. They find that there
weren’t any brakes at all; the driver had taken the brakes out so as to prevent,
even if they’d wanted to change their ways at the last moment, the car
possibly stopping. Then, through such a single case, the rule might get
changed. Now it’s the presumption that if there’s a railroad crossing accident
of a fairly general sort, it is suicide.

198

Part II

That matter may seem to be not terribly consequential in general, except
that there can be a whole set of circumstances where a rather general and
important moral rule of one sort or another gets handled in the same fashion.
The import of certain kinds of crimes, when they get publicized to some
extent, is that persons now begin to talk of the outcome of that trial
implicating how it is that the legal system works. So people said about the
Chessman case during its course, that if he didn’t go to the gas chamber, then
everybody would lose respect for our legal system, we would be giving free
play to deviants, murderers, whatever else, and that everybody would take it
that they could no longer hold the rule that the legal system by-and-large
works correctly.

That means, if exemplary occurrences can be formed up out of a single
occurrence, and if they can have the consequence of changing some people’s
knowledge, where furthermore the new piece of knowledge will have exactly
the same form as the old - that is, it also will be protected from induction,
then matters are extremely tender. You don’t get a step-by-step modification
of something. It’s frozen, it shifts, and it’s frozen again. Then the focus on
single occurrences is not that they are an exception, or that they slightly
modify what you know. It can be an attempt to shift whatever anybody has
known previously.

Lecture 12

‘Tearing down;’ Non-translatable

categories

I’m going to finish up our consideration of this segment by dealing with some
of its more overall features. We could say sort of roughly that the segment
itself has a kind of introduction, which is that one sentence that they all put
together, then there’s an argument for a bit, then what may be a transitional
segment which is that three-utterance interchange, “Face it you’re a poor little
rich kid,” “Yes Mommy,” “Face the music,” and following that, what they
eventually call this ‘tearing down.’

Now ‘tearing down’ is a procedure that can be done on a large variety of
occasions and in a large variety of settings, and one of the things we want to
ask about it is, Does it have any special place in the therapy situation, for one,
and at this point in the therapy situation — whatever point this may be. The
way I’m going to approach that is to ask whether that procedure which they
call ‘tearing down’ has any interesting relationship to what it is that at least
the patient in a therapy situation takes to be going on, at least some of the
time. One reason we could have for asking that sort of question might be, we
might feel as I feel that this segment is a kind of sample presented to the
‘company’ or new candidate, as to what it is that happens here. And we want
to see in what way it might be considered to be a sample.

Here is a piece of a conversation which takes place between a staff member
of a psychiatric agency and a woman who has, as she takes it, very serious
troubles, but who is very hesitant to go back into therapy. There’s an attempt
to convince her to do that, but she offers a characterization of therapy as she
experienced it before. She says,

A: I stopped at a very bad time, I know I did.

B: Uh huh

A: But the job situation got moving and I I stopped right when I was

beginning— well I would have been better if I had no therapy than to
stop when I stopped.

B\ Yeah, sounds like it.

A: Cause 1 stopped right when I was looking at the whole gory mess.

B : Uh huh

A: Before it started, you know, before I started to clean it up a little.

B : Uh huh

A: Was like a surgeon getting down to the disease, you know, and all of

a sudden he opens you up and there’s disease. Wow.

199

200

Part II

B : Uh huh

A: He knows what it is and quits there, (laughs)

B: Yeah, I suppose that’s it.

A: (laughs)

B: That’s a good example.

She offers this image, then, of being cut open. And that image of first being
torn apart as a first part of a therapy procedure, is one that, at least patients
hold often, anyway. What, in part, it consists of is, apparently, taking
varieties of things that the patient says about whatever, and either explicitly
formulating for them, or in some way leading them to see, that some set of
terms characterize their circumstances, where the terms can be more or less
technical in one or another way. Like, in the case I just quoted, right at the
beginning the doctor says, “Tell me about your problem” and she says, “I’m
ashamed to. I don’t want you to tell me I’m infantile, because I know that.”

That obviously can stand in some fairly nice relationship to what transpires
in this segment, and her characterization of what happened to her might stand
in some very nice relationship to that ‘tearing down’ procedure. And many
analysts of initiation procedures in general, and the early stages of things like
being hospitalized in mental hospitals, going through therapy, etc., talk of
them as degredation, mortification, stripping of the personality, and the like;
any of which might well be seen as the same thing named by the phrase
‘tearing down.’ But there’s a really deep problem involved. I don’t know the
best way of presenting the problem involved in sentences that have the form,
“Don’t tear him down” or “I don’t want you calling me infantile,” but I’ll
try a couple of ways. They may not seem initially to be just in point, but hold
on to it for a bit.

At the beginning of a book called Excess and Restraint by an Australian
anthropologist, Ronald Bemdt, we find him doing what anthropologists
occasionally do, and that is, to formulate what it is that he takes it he was seen
as by the natives when he arrived at this place that he went to work in. This
is from page ix of the preface:

We were viewed as returning spirits of the dead who had forgotten the

tongue of our fathers and wanted to relearn it.

Grammatically, that’s perfectly good English. I want to make the case that it’s
an asemantic statement. That is, it is not meaningful in English, though it
appears to be.

The first part, “We were viewed as,” is fine, and there are lots of things
that can be stuck into that for which this first part is a frame. And that frame
to some extent controls what can be put into the second part, in the sense that
there are lots of things that could be used as second parts of sentences, that
don’t fit here, syntactically or otherwise - for example, that isn’t the kind of
thing you say about humans.

I suppose it has to be clear that the second part, from “returning spirits”

Lecture 12

201

on, is some kind of translation from another language into English, and, in
part, that it is a translation is a considerable part of its warrant for being said.
Such that, if Berndt stood up at an anthropological convention and said, “I,
Ronald Berndt, am a returning spirit from such-and-such,” that would be
good grounds for them to think he’s insane. That is to say, if this were
English, it could not be a potentially correct description of what he might be.
But there are apparently ways that, if you take words one by one and in some
combinations, some combination of them can be produced to look something
like a Members’ category. Those words could somehow go together, and be
recognized as ‘a something in some world.’

We have some way of handling such a statement because you can say,
“Well, the initial thing is that he’s a stranger, and that’s some way of
characterizing strangers that involves somehow making him some sort of
member, and that must be one category that these people either had or
devised for handling such persons.” But we would take it that, except as a
matter of syntax, none of the controls on correct use of that clause ‘returning
spirits . . . etc. ’ are found in ordinary English. We would not know when it’s
been said correctly about someone, what are the consequences of it being said,
what you can infer from its being employed, etc., etc. It would be a sentence,
then, if it had any distribution in English, which would be reporting what it
was that was taking place for these other people. And it isn’t a proposed
introduction into English, I take it, of some category that we’ve lacked and
now are enriched by having. So, for example, until the term ‘schizophrenia’
was invented, it could be said that we lacked that category, and that given it,
a body of experience, a set of observations, etc., could now be ordered, and
‘schizophrenia’ now stand as their name - and be used, furthermore, as other
disease names might get used. But what kind of experience this term
‘returning spirits . . . etc.’ orders, and what class it would be a part of is, at
least to me, quite obscure. And yet there’s a sense in which we do understand,
by some mapping, what sort of thing he is talking about. And that, of course,
is an extremely interesting kind of fact.

The sentence, then, is an odd one, and the introduced category is not, as we
might put it, generative. That is to say, it’s not like a Reader’s Digest word
which, now having learned it, there are lots of places where you can use it. But
there may be somehow in there, an exploitation of English syntax in making
the translation and building a sentence like that. And that is, in part, I
suppose, why anthropologists will frequently not do this, but will say, ‘We
were viewed as . . .’ and give you some phrase in the native language, and
then say explicitly, “What we have here is a gloss of something that might
mean the following sorts of things in English.” But they still use the native
term, because then you can’t use the term in English; that is it remains a term
from that other language.

I introduce this sort of consideration since I want eventually to make the
case that one of the core problems for psychoanalysis and other psychother-
apies, and the social sciences in general, concerns the relationship between the
categories that those disciplines set up and the categories that members of the

202

Part II

society otherwise use. The problem is that somehow members take it that
such categories - ‘manic depressive,’ etc. - are additions to a list of categories
that exist already, and can be used in just the same fashion that the old ones
are usable. They may be better, but they do not otherwise modify the
structure of the class, of which people come to be seen as members. However,
the professional constructing these new categories may take it that one major
task he has is to somehow build them so they are unusable in the way that the
categories he sees them as replacing were usable. That is, the professionals put
it as a programmatic task that they would like to have it that the statement
“You’re a manic depressive,” for example, would be nonsense in ordinary
English, i.e., unless said by a therapist.

When Freud set out to build a ‘scientific psychology’ as he put it - and that
program of his, to build a scientific psychology, was among the last things he
ever got translated into English - one of the tasks he felt he faced in the first
instance, was to deal with the fact that everybody considered themselves to be
an authority in psychology. He was not locating a domain which persons
could take it from the beginning they had no idea what it is you might say
about it. But it wasn’t, of course, only the case that they had lots of views,
which they took to be well warranted, on psychological matters, but those
views, and the categories they used, were not morally neutral. That is to say,
in part, that any time one of them was asserted as being so about somebody,
it was also the case that some assessment of their status was being made - as
a good citizen, or whatever else.

Now, in many views of what could be called the philosophy of science, it
is taken that technical findings stand in some position of replacement to lay
findings. One can often have a picture of the state of lay knowledge which
stands in a strong relationship to what scientific knowledge is supposed to
look like. That is to say, that it’s a set of items about any or every given topic,
where what happens is, when a new item is introduced, it can operate to
expunge an old item. That goes for correct scientific procedure, and it has
been taken to go for what is called lay knowledge, as well - with some delay
perhaps, but essentially the same kind of thing can properly take place.

That’s a kind of tricky argument if you’re dealing with the structure of lay
knowledge, although it can be perfectly well satisfactory if you’re dealing with
the structure of scientific knowledge, in that you can set up by fiat what the
structure of scientific knowledge is, and you might get a list of items, and you
might, then, at any time, remove an item as a new one comes in that seems
inconsistent with it. However, what the structure of lay knowledge is, would
obviously be a problem for some discipline. It’s not in the first instance
known. It clearly has been a subject of some consideration since the Greeks,
anyway. And to propose that it’s a list of items is to make a rather daring
assertion. But it’s one that has been made.

As a further problem, there is the fact that it seems if an item is now
introduced to the scientific corpus, then the controls on its use can presumably
be very well laid out, and whether it is correctly used can be something
insisted on and enforced. And furthermore, it can be presumed that if it’s

Lecture 12

203

incorrectly used, you will in fact have hell to pay; that is to say, consequences
will just show up if you misuse some item. Now it’s not clear, in the first
instance, what kinds of uses these items do have when we re talking about lay
knowledge. It’s by no means clear, either, that they’re only the same uses, or
that ‘error’ - if we can formulate that - has the same short-run or long-run
chance of showing up.

For considerations like the ones I’ve offered, there is some considerable
concern about who would have the right, at all, to use new scientific
knowledge, or techniques, or whatever. Because, for one, if ir became
something that anybody could use, then God only knows what might be done
with it. And the simplest way for it to become something that anybody might
use, is for it somehow to be seen as a member of the corpus of correct
knowledge, undifferentiated as to scientific or lay. Where it’s the fact that
scientists say so that permits anybody now to assert and use it - where,
however, they need make no special change in how they use any category, to
use the new one.

One thing Freud was further concerned about was to somehow prevent
persons from using his categories in just the fashion they used the ones they
had before. And there’s very good reason for that, which is that if, say, ‘manic
depressive’ was a replacement for some lay term like ‘cranky,’ then whatever
assessments that were made about somebody said to be ‘cranky’ could be
made about a person said to be 'manic depressive,’ and that someone who was
said to be manic depressive might hear it as a kind of attack.

Now, I take it that, probably genetically, in the first stage of therapy,
anyway, that’s what’s seen as going on. And the term ‘infantile,’ for example,
is treated as being either the same as the lay term ‘infantile,’ when someone
says to an apparent adult, “You’re behaving like an infant,” or as some sort
of technical term which is nonetheless usable as the other term is. And that is
something that, for example, let’s say a field like astronomy has avoided, and
nobody gets very horrified to hear about the generative stars. But it’s not been
done in such fields as psychiatry. And the fact that it’s not been done has been
turned into a tremendous attack on psychiatry, by somebody like Thomas
Szasz, who will invoke the fact that the categories are used evaluationally
elsewhere, to propose that the phenomenon is that of making an evaluation,
and that their isn’t anything but that sort of knowledge being used.

That is, I suppose, a strong indication of the failure to cut off the
incorporation of new information into lay hands. And it would be in that sort
of way that one can see how the so-called ‘same procedure’ can be seen as one
that ‘you call X and we call Y.’ You call it ‘diagnosis’ and we call it ‘tearing
down.’ And one can see how, then, these patients could be characterized as
giving their version of what it is that’s going to happen in these sessions. It
remains the case that if the discipline retains the notion of items of knowledge
as the things it is basically dealing with, then it’s going to have to face the
problem that the items get taken and used. One thing that can be done, of
course, is to attempt to place anything that could be called such an item into
such a structure that if you pull the item out, it’s senseless.

Lecture 13
‘Everyone has to lie ’

The transcript of lecture 13 is incomplete and full of gaps. What there is of it
deals with the phenomenon that Sacks called ‘ Everyone has to lie.’ It became a
paper of the same name, and can be found in M. Sanches and B. G. Blount (eds),
Sociocultural Dimensions of Language Use (New York; Academic Press,
1973), 37—80. For the materials in lecture form, see Spring 1967 , lectures 8
and 9.

204

Lecture 14
The Navy pilot

The transcript of lecture 14 is full of gaps. In its place are Sacks' notes on the
data he considered in lecture 14, a New York Times article entitled ‘A navy
pilot calls Vietnam duty peak of career. ’ The materials are taken from Sacks’
Research Notes, vol. 2a (1964—5), 170—99.

MCD materials 1

New York Times Sat. May 29, 1965
From a long article:

A Navy Pilot Calls Vietnam Duty Peak of Career

by Seth S. King

Special to the New York Times

ABOARD THE CORAL SEA, in the South China Sea, May 27

Commander Jack H. Harris, leader of the Attack Squadron 155 aboard
the Coral Sea was explaining how a carrier pilot feels about Vietnam: “I
certainly don’t like getting shot at. But this is the top of my flying career
and it’s important I should know about it. I really feel I’m fortunate to
get the opportunity after 2 1 years in the Navy of combat experience. I
need it to be a real professional.”

How did he feel about knowing that even with all the care he took in
aiming only at military targets someone was probably being killed by
his bombs?

‘‘I certainly don’t like the idea that I might be killing anybody,” he
replied. "But I don’t lose any sleep over it. You have to be impersonal
in this business. Over North Vietnam I condition myself to think that
I’m a military man being shot at by another military man like myself.”

1 First, we may see one crucial matter here; and that is that he takes it
that there are alternative ways that he and those he is dealing with (bombing,
being fired at by), may be categorically formulated.

1 ‘MCD’ stands for Membership Categorization Device.

205

206

Part II

2 Second, that the choice of the categorizations is relevant to the
assessments he might make of his activities; now this is so in a strong sense.
To wit; that he sees that the formulations have alternative imports, and it is
not just the case that the formulations have alternative imports, and that
further by some work he is able to choose that formulation which has an
import that is consistent with what it is that he is doing, and that permits him
to do it without important moral consequences (he is able to continue
‘without losing sleep’). This situation of alternatives is then one to be marked
as one where he takes it that there is not some exclusively appropriate or
required choice of devices.

3 It may be much noted that the choice of the device is not just to be
made by reference to what they may make of his actions under some scheme
like reference to their consequences or the like, but by reference to how, given
the use of the consistency rule to formulate his alter egos, how it provides for
the formulation of his actions. The availability then of making his categori-
zation decision in such a way that it routinely provides for a categorization of
his opponents, and by virtue of the mutual categorization then an assessment
of either’s actions is a crucial matter.

4 Notice how it is that categorization can be held in some abeyance or the
use of particular devices not be foreclosed while making reference to
populations, i.e., the use of ‘anybody’ and ‘someone.’

5 Notice also how the characteristics of the required impersonality are
delivered, via the use of ‘you’ in a usage we have been considering.

6 There are several matters re ‘you have to be;’ is it the case that if you
have to be then you can be; or is it the case that, that you have to be provides
the grounds for being.

7 Now the import of the categorization of the others as well as himself,
and of the particular device chosen, is in part that an ethic follows, where:
what is proper for him is proper for them, or, equally, what is proper for them
is proper for him; and also, that the device chosen is not merely one which
assigns a category to he and to them, but also it offers a category to them,
which they may also use in just the way he uses them, i.e., that with it they
may formulate what is happening just in the terms he does, and with the
same consequences. If they accept that offer then at least there are no
complaints to be offered on their part about the error of his ways, except if
he happens to violate the norms that, given the device used, are operative;
that is, the device used does not provide that there are no rules for the actions
of either.

It is of course to be noted about the given device that it delocalizes the
affiliation of the personnel; what is relevant in the first instance is not that we
are the military of the US and they of Vietnam, but that we are both military;
it is that we are differentially affiliated which provides for who we shoot at;
otherwise we are colleagues.

8 One of the things to be noted about ‘you have to be impersonal in this
business’ is that its indexical features, particularly the ‘this,’ have some special
relevance. The statement is perhaps only incidentally correct for the military;

Lecture 14

207

it is correct for other businesses as well. And that feature of it may be quite
crucial. For, given the fact that it is true for other businesses as well, that the
consequences of impersonality differ may be rather irrelevant; so, businessmen
may be impersonal in different ways, or psychiatrists or priests; but if the
statement is true for them, and they are in a business for which the sentence
is true, they may not be in a position to without recourse to their own
circumstances importantly claim that he is doing wrong; for he can answer
that the same holds for them with the difference being that their impersonality
doesn’t directly concern killing. The property of prepared, or general purpose
indexical statements may be a rather crucial one; cf. ‘That’s the problem with
society’ and others like it.

And we know that persons who are attacked for the peculiar defects of
their activities are often seeking to formulate matters such that they can
claim that the same sort of defect, if not the details of it, hold elsewhere,
for the attacker, and perhaps for matters quite generally. This use of a
general purpose indexical statement may be very helpful in such an
attempt. And note of course that if one can make such a claim, that it does
seem that one prevents the attack from having its intended power; there is
something akin to an unclean hands argument. The unclean hands argu-
ment is further made powerful by then attaching as a further class beyond
its scope the ‘ivory tower’ or ‘bleeding hearts’ group as further persons who
are in no position to talk, because they’ve never met a payroll. One needn’t
then have a single argument to handle every locus of attack; but several can
be combined; where the first part gathers a bunch together, and then on
behalf of all, one uses the second part (where one proposes that one is
defending those who are talking oneself from the first group when one
attacks the second.)

It may then be that the use of ‘this’ with its possible extension being greater
than ‘military’ serves to cut off attacks by anyone who is an incumbent in
some business for which the statement is also correct. The use of ‘you’ in this
statement has been examined in the lectures on the first page of the initiation
ceremony.

9 One can notice that the relevance of a device with which a consistency
rule usage gives: military-military; is of separate importance than ‘you have
to be impersonal in this business.’ For that remark might be made by a
professional gunman, and it would not serve as an answer, as an adequate
account. Or would it? Or is it relevant?

10 Is there some suggestion such as: If we are both military men,
shooting at each other as we ought, then not only is it the case that neither has
any position to complain about the proper military doings of the other, but
also what we do is in some sense just our business, and having behaved
properly, no one else is in position to complain either? Alternatively or in
combination there is the issue that having taken care of part of their side by
giving them a military man to behave as I do, he being their representative
involves them sufficiently; they have no recourse because their side is behaving
as ours is.

208

Part II

11 One wants of course to notice that the formulation ‘this business’ is
of interest not only for the ‘this’ but for the fact that the use of ‘business’ is
itself quite relevant, i.e., the work need not be so formulated.

12 The way that having dealt with the opposing military being referred
to the population has been sufficiently handled is via the representativity of
certain categories of other categories, i.e., of a team for a city, and of an army
for a nation.

13 The remarks about ‘this’ or that sentence’s general purpose character
may hold as well for ‘it’ and the sentence it is contained in, i.e., ‘I don’t lose
any sleep over it.’ And the combination may work together to get a general
defense which has to be accepted or the attack it succumbs to may be used in
too many other places.

14 I take it that it is quite important that the character of the wrongness
or possible wrongness of what is being done is not formulated until both the
doer and the done to are categorized in some other fashion than ‘someone;’
though it is apparently the case that the possible wrongness is present once it
is established that what is going on, killing, is being done to someone. Thus
the someone fact sets up a problem of formulating an account; but given only
the someone formulation, the fact that an account can be made is left open,
i.e., it is not foreclosed that it can not be made.

Here of course one has to consider the possibility that there are ethics for
which it would be sufficient to establish the wrongness if one had found or
possibly found no more than that it was a case of someone being killed.
Weber on the difference between India and the West.

15 Note also in this regard that ‘someone’ is sufficient to set up the
problem and to deal with it via the categorization and that one need not make
a count of how many they are. On the one hand, that is, that one need not
know, to make the claim that an account is appropriate, how many in fact;
one may suppose that some, or may be assured that some, without knowing
how many, and also the account that is constructed is not apparently affected
by the size of that some. Though it is the case that if the size is known, that
size can be relevant to the need for and the character of an acceptable
account.

16 Re earlier remarks; if it were the case that, that you had to be
impersonal in this business held only for this business, then it might be that
doing this business would be wrong in the first instance, or could be. That it
isn’t so only for this business is crucial to the usability of the sentence ‘in this
business. ’

17 Re impersonality; it is not simply that in doing the activities this is
something one is forced to do, but in the formulation of the account what has
been done is that an account has been offered which could be offered by any
person involved on either side of the affair; and via that fact the impersonality
point gets perhaps a further emphasis; thus having been asked; how do you?
(deal with it), there is on the one hand a claim of need for impersonalizing and
then an account which is one that is formulated in such terms that it could be
offered by anyone involved, i.e., is impersonalized. That is in addition to the

Lecture 14 209

point that the remark ‘you have to be impersonal in this business’ is a sort of
ideal impersonal remark.

18 ‘Impersonality’ as a requirement is of course not an unequivocal
object; for it in some circumstances could not serve as part of an excuse, but
might be offered as a difficulty, i.e., I wish I could know who it was that I was
(killing, saving, etc.) so that I could get special pleasure from that; as, e.g., in
certain charities the possible impersonality is handled by one being promised
mail from those one feeds, and then some relation is established between
feeders and fed of a personal sort. It is not that one has to kill impersonally
that it is a difficulty; so that it would be nice if one could kill personally, but
that one can’t think of killing, or be concerned with that, even if it is in the
first instance the case that one does not know who one kills; where one has a
possible equivocal that comes on with perfect unequivocality that is of some
special import. But maybe no such sense of impersonal is involved; rather,
that one has not to consider that one is killing persons.

19 ‘Over North Vietnam I condition myself;’ this pair of phrases is quite
similar to; ‘99 percent of the time a guy could pull up to me;’ in both cases
the first phrase which modifies some later phrase does not modify the one that
consecutively follows it. It clearly could be placed before the phrase it modifies,
but it isn’t. Clearly enough it can do what it needs anyway, if ‘anyway’ is
relevant; but the positioning freedom is something to be noted if the grounds
for the use of the freedom are not yet apparent. (The second quote is to be
found in the initiation ceremony data, i.e., group therapy 6 pages.)

20 The clarity of the fact that his account not only holds for him and for
any others on his side, but for the ones on the other side is provided for by the
fact that what he states is the account that might be offered by one who
subscribed to his formulation and who was a Vietnamese military man, i.e.,
the one speaking formulates, not his bombing as the relevant exemplary
actions but his being shot at, i.e., he explicitly is offering an account, i.e., the
fact that he is shot at as something he excuses because the other guy is a
military man shooting at him, a military man (providing the hope that the
other will excuse him as he excuses the other).

2 1 Query; how can he come to say ‘the idea that I might be’ as, cf. ‘the
fact that I am.’ One possibility which I want to explore is that there is a kind
of transitivity to an indefiniteness feature. Perhaps it is the case that if it is not
observably definite that the given actor has killed some particular individual
(by which I don’t mean someone whose name, e.g., he knows, but that he sees
that someone in fact has been made dead) then even though it may be quite
clear that the bunch of men have killed a bunch of people - and that they
would have failed, for example, if they had not - the non-locatability of flier
A killing some person A may permit it being seen as indefinite that he has
killed someone; if it is indefinite that A has killed some person, then it may
be indefinite that A has killed any person; and if that is so then it may be that
he can talk of ‘might’ and of the idea, as cf. the fact.

Now the point of this is not just with regard to killing, but might have an
import in, say, market events, where although it is perfectly clear that the

210

Part II

collection of businessmen have done something that has as its consequence
that some collection of others have been injured, the non-locability for each
businessman of some person that they have injured may permit indefiniteness
to take a transitive relational form, and also to be operative as ‘an idea’ that
one may have to come to terms with; where, that it is an idea that one has to
come to terms with, provides a variety of ways that one can use: the ways
available for coming to terms with ideas that are unpleasant.

The phenomenon of organizational, institutionalized destruction, cf. Nazi
operations, etc. are obviously possibly involved here (Note that one possible,
and terribly interesting correlate involves the fact that the ‘coming to terms
with the idea’ can be an immensely painful thing that persons who are ‘not
directly involved’ may feel; there can be, then, a generalizing import to the
situation as well; so those involved directly may have the fact of transitivity
for indefiniteness and thereby the issue of ‘the idea’ being a real gain to their
circumstances; while those not involved directly can -find the idea an immense
source of pain. Much current and classical philosophy, theology here.)

In this regard note that it seems that it is the case that to the remark ‘There
are many things wrong with this,’ the following is a good answer: ‘Name
one,’ where, that one cannot be named or is not named is seen to undercut the
claim of ‘many;’ alternatively if one is named then that may have more
import than that one has been named; that interchange may also get its
character via the transitivity of indefiniteness.

It might be said that having said transitivity we want to be saying that the
matter is two directional, i.e., not only if one is indefinite then many. That
may be so; have to see.

22 It is of course to be noted that there is not any ‘necessary’ relation
between the possible transitivity relation and that fact that given its operation
one must consider the idea, or that one can consider the idea and not the fact.
That is partially separate; but it may well also hold that if definite then one
must consider the fact; if indefinite then one can consider the idea as cf. the
fact (which does not mean the non-relevance of possibility, but: relevance of
possibility but non-definiteness of it.) Then, if indefinite, one must consider
either (idea, fact) and can consider one of them.

23 We are of course supposing that at least with regard to the
combination (idea or fact + might or am) that the choice, whether required or
not, of a first, operates to control the choice of a second, and that at least if
‘fact’ is chosen then ‘might’ is excluded as a PCD; without however
suggesting that if ‘idea’ is chosen ‘am’ is excluded. Now we have of course to
determine that in fact there are some controls on combination here.

24 The first part; ‘I certainly don’t like the idea that’ is not optional if the
position one thereafter takes is to get credit; one can’t properly say (with it to
have the same reception) either ‘I like the idea,’ or ‘I don’t mind the idea’ or
‘I never thought about it’ or leave the space blank - there is a space in the
position assertion for that preface; ‘I don’t lose any sleep over it’ is not just
incomplete, but it is understood as having ‘I don’t mind’ as its prepart. Thus
there is what is literally a formula; ‘I certainly don’t like the idea that I might

Lecture 14

211

be doing X, but I don’t lose any sleep over it;’ i.e., + some account which
is signalled as coming, by the use of ‘but.’ In this context the ‘but clause’
serves to provide that the utterance is still open, and that furthermore
something that is proposedly an account of that ‘but clause’ will be offered.
You do X (X is something which without it being said is something
problematic, i.e., presumptively wrong); I don’t like doing it, but I can,
because. This package is a general purpose one and can be used quite widely.
In GT, a guy says about his girlfriend, “she’s on the girls tumbling team. I
mean, she doesn’t like it, but she needs the credits’’ (this may not be exact;
the quotes are in the data somewhere). 2 When the formula is present it is for
the analyst a strong way of identifying a low valued object, i.e., that thing
which is the X in the formula’s first part; there is of course also the account
part. [Then, handwritten :] The formula is crucial not only ‘because’ these are
low-valued, but also because they can be done given the formula’s use.

25 The first part of the formula, ‘I certainly don’t like,’ acknowledges
the fact that the question, ‘How do you feel about. . .?’ operates on an object
which has some feeling about it as its proper one. Thus, the very question
involves a tapping into an evaluational schema, and the formulaic return
acknowledges not only that, but that the question not only taps in, but
indicates that the user is committed to that scheme or recognizes its
legitimacy, and that the answerer knows the schema, knows that it generates
the question, knows further that there is a proper position to be held, but that
having subscribed to the position, a further statement is not foreclosed. And
it is that subscription to the position’s status that is crucial. Having made it,
one can account for the action in an acceptable way. (Note: I am not offering
this statement as just about the given piece of data, as ought to be obvious.)

A return such as ‘Why do you ask?’ would be immensely informative
about the subscription the doer of the return has. As perhaps the need to lay
out in the question, that some hold that the thing being done is wrong; that
is not a stronger, but a weaker formulation than the simple question, ‘How
do you feel about . . . ?’

26 Notice that the term ‘like’ is not additive of features, in the phrase
‘another military man like myself.’ What it is is an affirmation of the
consequence of his being a military man (given the fact that I am one). It is
an equivalence signalling term, which points out the relevance of the item it
follows, i.e., I mention military man because by that I establish the
equivalence of this other guy and me, which equivalence is establishable in
that case sufficiently by our common locus in that category.

27 It is perhaps the case that the phrase ‘another military man like
myself is handleable in part via the remarks on transitivity relation, i.e., if he
is a military man, then he is like any other, and being like any other, he is like
myself. 3 (It is not directly a class inclusion argument; though the middle is

2 Actual quote: “She’s in the girls’ tumbling team, (pause) I mean she doesn’t like
it she’s just on it because she needs the credits.”

3 Written in: “No: this is a symmetry issue: will consider it in a bit.”

212

Part II

he - class - me; given this format it is obvious that me - class - he is just as
usable; the use of the former is tactically nice; for it is me who is being shot
at in the first instance. Perhaps then forget the transitivity point here.)

28 ‘Idea’ may be not only sequentially related to ‘might’ but also to
think in the next sentence.

29 Notice that there is a nice relation between ‘how do you feel about’
and 'I don’t lose any sleep over it.’ The latter is of course an idiom answer;
but it is not simply a phrase that has happened to have some standard
meaning. In our analysis of ‘How are you feeling?’ we noted that the
specification class of categories that are relevant given the choice of an answer
from the negative subset has as one of its members (the PSW class) ‘sleep.’ 4

Given that locus of ‘sleep’ and the conditions of its relevance the idiom is
not a merely accidental freezing. (The discussion is in the first instance to be
found in the Diagnosis of Depression drafts.) The use of the idiom then may
provide in the first instance that there is some relevance to the fact, as I have
asserted, or as may be supposed that I am in a negative subset on ‘feeling’
whether I’ve said so or not.

What has happened is of course a kind of generalization; where the
addition of about + whatever it precedes, does not affect the structure of the
relation between ‘feeling’ and the consequences of the negative subset.
Retaining then the private imagery or relevancies; as in ‘Johnson makes me
nauseous.’ The test of something’s consequentiality, by the way, is under
some formulations to be made by reference to whether such intimate
formulations of consequences, or such intimate consequences are in fact made
to initially public matters. If they are, then it’s consequential; if not, then not.
Cf. Kennedy’s death was like something happening to a member of the
family; the war in Vietnam is consequential now because it is affecting
millions of Americans directly, vis-a-vis their sons, etc. Now what we want to
notice is not so much that or whether such assertions are correct or not, but
that they are the ways of making the point of ‘effect’ strongestly. And that is
presumably a culturally relative fact. (Though some theorists, Weber, for one,
I think, might argue, or did argue that things had in the end to get posed in
this format if they were to count practically.)

MCD (1)

We have not considered whether there are any controls on the sequence a
population ought to be categorized in, or whether and how that might matter.
Now it can be made clear that it could matter; there is the one person
problem; and there is the nth person solution; 5 so, even for two persons order
of categorizing might matter.

4 See Spring 1967, lectures 8 and 9, on the phenomenon “Everyone has to lie.”
A ‘PSW’ category is a ‘Personal State, Which-Type’ category, such as sleep or
appetite.

5 See Spring 1966, lecture 7, pp. 325-7.

Lecture 14

213

What sorts of controls are there on the sequence for doing categorization;
it may be noted, though the matter is not directly in point, or clearly so, that
it may regularly be the case that while there is some freedom in choosing the
first person to introduce someone to, thereafter the order may be quite
fixed.

To see right off how it can matter, there are certain categories which are
clearly not for a population of only one person, and which are first-categorized
categories, i.e., their use involves the person categorized by reference to them
being the first one categorized in some group that is larger than or equal to
two. For example, the categories, murderer, attacker, rapist. And there are
some second categories, i.e., whose use indicates that the person being so
categorized was not categorized first, e.g., self-defense murderer . . .

These few facts may have some consequences; first of all the set of
categories to be used on some population, i.e., those to be chosen from those
that might be chosen, may not be independent of the order in which persons
must be categorized, or may be crucial to what order in fact is used. The fact
that there is some freedom in sequence of categorizing may be quite useful in
expanding the set of categories that can readily be used, or effectively be used.
At least this could be true for those devices which contain first-categorized
categories and second-categorized categories.

MCD (2)

1 Sequence of categorizing may be relevant for which device as well as
for which categories.

2 There are some categories which obviously are ‘first categorized
persons’ categories, e.g., ‘winner’ (win, place, show). It is not always
necessary that for some categories which are in one way or another ordered,
that the first of the order be the first categorized, or the first category be
applied first, but it is regularly done, e.g., regularly the announcement of
scores or other contest outcomes are announced winner first. Cf. here,
Malkiel on ordered binomials. For some, however, order of categorizing
personnel is crucial to use of categories e.g., attacker-attacked. Of course,
first, second . . . Sometimes announcing is done in what is clearly seen as
reverse order; though determining is done in proper order; announcing for
‘suspense.’

3 The notion ‘impersonal’ is used as a term which is, it seems, not only
one of some collection, perhaps the following: (personal, impersonal), but
where the set is evaluationally ordered and has a given sense for the objects it
applies to which are also evaluational. To be impersonal is not good; it
requires an account.

Furthermore the activity characterized by the term impersonal is not good.
The latter fact is quite important, for it informs us on the state of a kind of
battle. There are proper referents of ‘impersonal’ such that subscription to a

214

Part II

moral order involves not only giving an account for what it is one does
impersonally, but also that certain things are to be seen as done personally is
not to provide that thereby no account is required, but to indicate that one is
not, in some important ways, a member. Note too; for some matters there is
not an in-out group morality. Killing anyone is wrong, accountable, and may
require reference to impersonality, or personality. Not just wrong when done
to members.

MCD (3)

Personal (love, hate)

Impersonal - neither, business.

If a killer were personal then he would have to hate; now, to have a way
of doing killing without hate requires some formulation for which hate is
irrelevant; hence, business.

It is apparently possible to do things to persons for which it is irrelevant
either or both that they are persons or that you are one. One does something
for which in the first place emotions are involved, on your part. The fact that
emotions are involved, i.e., that they are bound to activities, is, in the first
instance, crucial enough; that the collection of emotions can be substituted for
is another order of fact. It means that one need not, retaining the former,
arrange matters so that ‘hate’ is held. Alternatively it means that modification
need not be done so that the boundedness of the activity is modified; that one
kills, sometimes out of love (euthanasia, sadism). But that the collection is
irrelevant.

Now what is to be seen is that this alternative is not made for some
particular activity, but is available for a great many, and not only those for
which hate is otherwise appropriately the basis; but those for which love is,
too. One wants to see that access of persons to ‘impersonality’ is restricted; to
claim in some instances that one felt nothing when doing a killing is to
provide the warrant for determinations of psychopathology.

The fact that such a statement has a wide scope of ‘businesses’ for which
it is applicable and that even for the ‘military’ it has not been invented by this
fellow, or by some one for this war, is a most criterial sort of fact. It is for one
the sort of fact one might point to in desiring to warrant the claim that there
is not a strong tie between a particular basis for a conflict and the fact of
conflict, i.e., for treating ideologies as rationalizations.

If the scope of ‘this’ provides a set of objects that have some equivalence,
that is one thing; but if further there is an equivalence established between the
things done in those businesses that is another, quite another.

MCD (4)

The fact that the statement has the ‘you’ subject which here seems to involve
in part that the doings concerned are doings of particular people, though not

Lecture 14

215

necessarily just one of them. Now it sometimes is possible to avoid the moral
issues by so arranging things that consequences are not relevantly focused on
by any or by most persons involved in their production. The jet pilot may see
his particular involvement whereas thousands of others whose actions make
his work possible do not.

The relevance of emotions:

1 They are an account by virtue of being a cause; and as such, they can be
an adequate cause and not be an adequate account. Both love and hate can
fail as adequate accounts while being perfectly adequate causes.

2 They are proper causes in a different way than they are or can be proper
accounts.

3 For them to be both, something more is needed than their causal state,
and if something more is needed, then that which makes an adequate
account can be present even if they are not.

Personal— Impersonal (1)

With regard to the issue of the personal or impersonal character of his feelings
and of the need to be impersonal in one’s feelings, one quite important basis
for this need may be gotten at by considering first the heading to the story,
‘Navy Pilot . . .’ He is a navy pilot, where that he is one of some bunch is
quite relevant to what it is that is to be done with, and the sense of, his answer.
For the answer is properly to be seen as representative, and as such it may be
relevant that he offers an answer which is not only in fact representative, but
is one that others can live with.

Now being engaged in doing that sort of thing, that is, where one’s answer
as an account of one’s action is also and at the same time a thing which can
stand as a position of others who do what you do, or pose a problem for them
to which they may have to come to terms, involves then a kind of
responsibility in posing that question and also a kind of need to be impersonal
in making the answer; impersonal in the sense of being on behalf of some
unspecified set of others. A personal answer is one that might not commit the
others, to which, that they feel that way is to be determined by inquiring of
each of them; an answer formulated as a pilot, in that impersonality, may be
one that, having that character, can be assumed to be how such as are so
situated feel, where it can be assumed that they feel that way, unless someone
happens to disaffiliate himself, or to offer how he feels personally, i.e., not as
a pilot.

In this regard, one interesting consequence is that how it is that nobody in
particular, not any of them in fact, feel, as persons may be presented, each
action on behalf of all offer what it is that they take it such as they ought to
be feeling. That is some of the kind of problem that is posed by reference to
how you feel as a pilot and how you feel personally, where the latter may be
treated as how you really feel. It is in perhaps consideration of this sort of
matter that one can most readily come to see how it is that small samples can

216

Part II

work out, i.e., that asking one or two gets what one would get were one to
ask many, where one asks of them to answer in the guise of their status as a
member of a category of which there are both many and each a representative.
To get their ‘personal’ feeling is to get them to formulate matters via quite
different structures, e.g. family stuff. Cf. Chessman.

Personal-Impersonal (2)

In our first point we say that he takes it that there is an alternative to
‘impersonal.’ What is our warrant for such an assertion?

1 I say that he takes it that there are alternatives and that impersonal is
not therefore all that is available, not simply by virtue of the assertion that if
impersonal is seen as part of a device then there are other devices than the one
it is a part of, but because the device of which it is a part or seems to be a part
contains at least another category, and that other is an alternative category -
that is, personal. Impersonal as a member of the device is clearly an alternative
to that other. (That he knows other devices is evident from, e.g., his reference
to a ‘wife;’ where he might refer to himself when flying, or otherwise, as
‘husband’.)

His wife, Sue, lives in San Diego with their two children. “If you are a

career Navy man, you have to have a career-type wife and I’m fortunate

in having one,” the Commander said.

2 It is also the case that he uses ‘have to’ in a restricted necessity sense,
which excludes the use of, but not the presence of alternatives. He does so
clearly in another part of his remarks, and may then be doing so in this part.
That is, he does so when referring to the professional-type wife, where that is
asserted via ‘have to have,’ and it is added that he is ‘fortunate to have one,’
indicating that one might have to have and yet not have.

3 Furthermore our proposal that ‘personal’ is an alternative and that
therefore he employs impersonal as an alternative, is warranted by the
negative precedent to the impersonal statement, ‘I don’t lose any sleep over
it,’ where the term used there is an idiomatic reference to a consequence of
personal involvement; one that has its character as idiomatic reference by
virtue of the status of ‘sleep’ as a personal state, and the loss of it as something
which is accounted for by personal problems. One indicates that something is
not affecting one, or not affecting one personally, by reference to its not having
this sort of consequence where, the thing seen as personal, it would properly
and expectably have such a consequence - were he to see himself as
murdering people he ought to lose sleep over it.

Note: ‘being killed by his bombs’ (a paraphrase transform of ‘being
killed by your bombs’) is by the pilot transformed explicitly into ‘I might
be killing anybody;’ he does not retain the format as, e.g., ‘my bombs
might be killing anybody.’ While there are ways that that might be

Lecture 14

217

preferable and permit a formulation of matters that could be pleasant, there
are strong purposes for which the transformation is crucial and the proper
use - the alternative being polite or an offered out which is rejectable or not
accepted - and that is, for purposes of scoring his successes, how many
planes he’s shot down, railroad cars destroyed, and the like. For these uses
then ‘his bombs’ are ‘him.’ Whether it might be possible to set up a
situation where it was he that destroyed the military target and his bombs
that killed others is something that might be considered; whether if one
chooses to formulate the former one must accept the latter as well by some
consistency notion is another issue, related, whose scope of operation needs
be considered.

An Answer to a Question (1)

Having offered in the draft paper on ‘everyone has to lie’ an analysis directed
to showing how it is that a statement a member makes might be said to be
true, we might consider a kind of parallel analysis; we might engage in
attempting to determine, considering the pilot interview excerpt we are
focussing on, how it is that his remarks might be said to be an answer to the
question he is addressed. What is it that we have to construct so as to provide
that he has done an answer, if we can so construct an apparatus? We might
consider the matter noticing that narrowly, or for coding purposes anyway, it
might not seem that he has given an answer; having been asked how he feels
about possibly killing somebody or anybody he replies that he doesn’t like it
but puts his main, or more extended emphasis on how he feels about being
shot at, and how he can excuse his attackers. The latter part of his remarks
are clearly relevant to whether he has a complete answer, or a possibly
adequate one, and in examining them we might come up with matters
relevant to what it is that an answer, or saying that there has been an
answer, involves.

An Answer to a Question (2)

In regard to the issue of how his remarks might be an answer, the first thing
we might consider is how it is that the utterance it is directed to is a question,
and what sort of a question it is.

1 As a question it is clearly one that is highly informed; it is posed by
one who has a rather considerable store of knowledge about relevant
matters.

2 It is also an overtly sympathetic question; by that we mean to point to
the sorts of considerations it imputes to the pilot, considerations which the
pilot has not asserted he has, and which the reporter in any event would not
know definitively that he has, e.g., that he aims carefully only at military
targets. The considerations are of course imputable by reference to what it is
the pilot ought to be doing, i.e., the regulations under which he properly

218

Part II

proceeds. Thus we say that there is an informed basis to the question since the
regulations that properly govern the pilot are employed in forming the
question he has been asked.

3 However there is a way in which it might not be what it is overtly, i.e.,
sympathetic, for it proceeds, those things aside, nevertheless you are doing
killing of persons who might not be considered as military targets aren’t you,
and how do you deal with that. Whereas had it been less overtly sympathetic,
i.e., had asked how do you feel about killing people with your bombs, an
answer might have been made that I try very hard not to kill people but to
bomb only military targets; then it would be necessary to amend the first
question in making another, and the fact that the position was overtly hostile
might be used to doubt the seriousness of the question. Its possible
sympathetic character is then of equivocal import.

4 The question furthermore is informed in a deeper way; it seems to
indicate and is seen as indicating in its very formulation that the fellow being
asked does not like this fact that he may be, is, killing people, it is indicating
this by setting the possibility up by reference to the care that would in the first
instance be taken not to have that consequence more than is necessary, or than
chance would require. That is, were it not the case that the questioner took it
that the fellow did not like or would not like the killing he needn’t have set
up the conditions as ‘even with all the care he took not to.’ The latter phrase
does not merely propose the questioner’s attitude, but proposes that the
questioner knows the basic attitude and its consequences for the answerer, and
that he accepts these.

Thus again the sympathetic character of the question is presented. Where
that is sympathetic in that the question is taking it that what properly ought
to be felt is felt by this fellow, rather than requiring of him that he assert what
he properly feels, or that what is proper is something that is a question which
we shall see if he knows the answer to. In this way of course there is some
informing to the answerer of the sorts of things he might in part deal with in
the question; not only what attitude he ought to express, but what he need
not say, that information being contained in the question, where both parts
will presumably be available to those who encounter the answer, i.e., in the
newspaper report; where in that environment the question and answer will
together contain the information that the reader will get, i.e., that it is so that
he aims carefully at military targets, and that he does so by virtue in part of
the fact that he doesn’t like to kill people. The question could have proceeded,
‘as a professional you really don’t mind killing people do you?’ or, ‘as a
professional do you have any attitude at all towards killing people?’ and the
like. He is not that sort of professional the question suggests, and the answer
acknowledges.

An Answer to a Question (3)

It is perhaps overly weak to just suggest that the question may be formulated

Lecture 14

219

in a rather sympathetic way; in such a way as to perhaps suggest or to suppose
that there is indeed an adequate answer to a question that might seem to lack
one. That is, it is a question which is perhaps not so much torturing the
answerer as to how to find an answer to it, it being something he has to
grapple with, but the sort of question to which a good answer may not be
known by outsiders, the public, and which is asked so as to give them a good
answer, or to indicate that the soldiers have indeed dealt with it and answered
it.

In a stronger way, then, perhaps it is the case that the very asking of the
question, or at least the reporting of its having been asked and answered, is
done by virtue of the fact that a good answer’s existence is known to the
questioner, but it is useful that the answer be something that the ones most
closely involved are seen to have; then it gets the character not of just a good
answer but of one that is operative fact. (There could be good answers which
would not be operative were they not known and used by those who were
specially relevant.)

This fact that if a question could be quite embarrassing to have to answer,
then someone sympathetic or someone fearful would not ask it were they not
in some way assured that a good answer could be made is of course quite an
old observation. The reader may be referred to Strauss’ discussion early in On
Tyranny, which I shall insert in any event in a while. 6 The fact, then, that the
question is asked at all or reported to have been asked may be quite relevant
to guessing the sort of answer that is offered, i.e., that it is something that the
asker takes as adequate, and takes the other to take as adequate.

(The following from Thucydides, book one)

For in these early times, as communication by sea became easier, so
piracy became a common profession both among the Hellenes and
among the Barbarians who lived on the coast and in the islands ... At
this time such a profession, so far from being regarded as disgraceful,
was considered quite honourable. It is an attitude that can be illustrated
even today by some of the inhabitants of the mainland among whom
successful piracy is regarded as something to be proud of; and in the old
poets, too, we find that the regular question always asked of those who
arrive by sea is “Are you pirates?’’ It is never assumed either that those
who were so questioned would shrink from admitting the fact, or that
those who were interested in finding out the fact would reproach them
with it.

‘But I don't lose any sleep over it’

The item selected and reported on here idiomatically - the sentence is an
idiom - is worth some attention:

6 See pp. 22 1-2.

220

Fart II

1 It is heard and is presumably produced as a representative item. We do
not take it that having heard him say that, what we ought to do or what the
interviewer ought to do is to probe on further matters from the collection of
which ‘sleep’ is a member - to ask, e.g., has your appetite suffered over it?
The item, then, seems to serve as a means of indicating that on the various
items of the class of which this one is a member, there are no consequences
of the thing that is being considered as one that might have such
consequences.

2 The relevance of the representative item is furthermore clear; that he
doesn’t lose any sleep over it means that it is not having personal consequences
of a negative sort for him. He needn’t interpret his answer.

It may be noted that the understandability of this import of his not losing
sleep doesn’t depend on any intimate knowledge of his sleep patterns, of, e.g.,
whether effects on his sleep are the usual way that troublesome actions of his
show up, or whether there is perhaps some better indicator for him - perhaps
he never loses sleep over anything, but he breaks out in a rash. The use of the
item as representative for him, is effective, then, by virtue of the fact that it
is an item which, unless otherwise specified, is representative for anyone, and
in the same way, i.e., as a negative indicator.

3 Furthermore, just as it is the case that it is apparently quite under-
standable to the interviewer what status the item has for this fellow whom he
barely knows, so, too, is it the case that the interviewer can take it that the
fellow is, unless otherwise specified, an authority on the causes of his losses of
sleep. This matter, that he would know why he has lost sleep if he had, is of
equal relevance to the preceding ones.

4 The idiom is of some interest in a way similar to the other one he
uses - ‘you have to be impersonal in this business,’ where ‘it’ seems to have
a similar character to ‘this;’ the idiom is pre-constructed for some class of
matters to which it may apply, where it is not just that sleep is representative
of the way that personal consequences take effect, but that each of the items
of the class that we have personal consequences, may be expected to have their
consequence via the representative personal consequences. It is not then the
case that some items that have personal consequences have one sort of
representative effect, and others others, or even if it is, the representative items
can be employed to present the fact of an effect or its lack. One knows, that
is, that if an item is one that may have personal consequences, even though
one has not done the thing before, what sorts of personal consequences one
should look to, to see whether it is having those effects.

This is quite crucial; for it involves, then, the fact that one can be
monitoring for the consequences of matters that one has not encountered; and
even then, if one does not know what is causing them, one can know that
something is wrong, there is some unknown cause, by virtue of the fact that
some things are happening which are properly formulated as ‘consequences’
of some personal trouble. Thus a search procedure may be set up to find the
item. And a decision may be made that it is this or that, or not. In this way,
the fact that one starts with a limited set of personal consequences, or things

Lecture 14

221

that are properly seen as such, means that the set of things that can cause them
need only be known as to its name and need not have each item in the first
instance known.

[Written in:) I have often talked of items as ‘representative members’ of
classes. With the above you can get a very clear idea of what is involved in
that.

The fact that he is an authority on what the sorts of things are which cause
him to lose sleep, and in a strong way that he can isolate that thing which on
some occasion has caused him to lose sleep is obviously quite a fundamental
fact. But it must be considered in another light as well; and that is that
interviewer whom it might be said barely knows him, also knows what sort
of thing might do this, and is in some position to decide whether the given
item would do it. Now those two facts are terribly relevant to each other. If
we were to say that the fact that he is an authority, and furthermore ought to
be an authority, on what it is that causes him to lose sleep, is a partial answer
to the question, or a partial criterion of ‘knowing oneself,’ then we can see the
sort of thing that that maxim meant for the Greeks. Knowing thyself does not
mean knowing something very private, it means knowing oneself as a
member of a community, knowing, that is, the things that obtain for one,
which obtain for persons commonly. Heraclitus’ fragments and Socrates’
certainly had such an intention; and they did not involve knowing something
about one which was distinctive, or special, or private.

The moral order provides one with the things not only which properly
affect, control one’s behavior, but with the materials in terms of which one can
come to know oneself, and in doing so, know anyone else, and in doing so,
be accessible to others. Thence, the hopes that may be attached to the phrase,
or the program, ‘know thyself.’

An excerpt from On Tyranny can be found in volume 2 of Sacks' Research
Notes (pp. 75 7—8). It does not, however, seem to bear on the issue of sympathetic
questions vis-a-vis the availability of a good answer. Rather it in part seems to
bear on the issue of ‘ tearing down' , considered in lecture 12 (p. 200):

On Tyranny by Leo Strauss (New York, 1948), pp. 18-19-

To express this more generally, Simonides knew that Hiero was not a
perfect ruler. He decided to teach him how to rule well as a tyrant. More
specifically, he considered it advisable to warn him against certain grave
mistakes. But, to say nothing of common politeness, no one wishes to
rebuke, or to speak against, a tyrant in his presence. Simonides had, then,
by the least offensive means to reduce the tyrant to a mood in which the
latter would be pleased to listen attentively to, and even to ask for, the
poet’s advice. He had at the same time, or by the same action, to
convince Hiero of his competence to give sound advice to a tyrant.

Before Simonides can teach Hiero how to rule as a tyrant, he has to
make him aware, or to remind him, of the difficulties with which he is

222

Part II

beset and which he cannot overcome, of the shortcomings of his rule,
and indeed of his whole life. To be made aware by someone else of one’s
own shortcomings means, for most people, to be humbled by the
censor. Simonides has to humble the tyrant; he has to reduce him to a
condition of inferiority; or, to describe Simonides’ intention in the light
of the aim apparently achieved by him, he has to dishearten the tyrant.
Moreover, if he intends to use Hiero’s recognition of his shortcomings as
the starting point for his teaching, he has to induce Hiero expressly to
grant all the relevant unpleasant facts about his life. The least he can do,
in order to avoid unncessary offense, is to talk, not about Hiero’s life,
but about a more general, a less offensive, subject. To begin with, we
shall assume that when starting a conversation with Hiero about the
relative desirability of the life of the tyrant and private life, he is guided
by the intention to dishearten the tyrant by a comparison of the life of
the tyrant, and therewith of Hiero’s own life, with private life.

To reach this immediate aim in the least offensive manner, Simonides
has to create a situation in which, not he, but the tyrant himself,
explains the shortcomings of his life, or of tyrannical life in general, and
a situation in which, moreover, the tyrant does this normally unpleasant
work, not only spontaneously, but even gladly. The artifice by means of
which Simonides brings about this result consists in his giving to Hiero
an opportunity of vindicating his superiority while demonstrating his
inferiority. He starts the conversation by presenting himself explicitly as
a man who has to learn from Hiero, or who is, in a certain respect, less
wise than Hiero, or by assuming the role of the pupil. Thereafter, he
makes himself the spokesman of the opinion that tyrannical life is more
desirable than private life; i.e., of the crude opinion about tyranny which
is characteristic of the unwise, of the multitude, or the vulgar. He thus
presents himself tacitly, and therefore all the more effectively, as a man
who is absolutely less wise than Hiero. He thus tempts Hiero to assume
the role of the teacher. He succeeds in seducing him into refuting the
vulgar opinion, and thus into proving that tyrannical life, and hence his
own life, is extremely unhappy. Hiero vindicates his superiority by
winning his argument, which as far as its content is concerned, would be
merely depressing for him: by proving that he is extremely unhappy, he
proves that he is wiser than the wise Simonides. Yet his victory is his
defeat. By appealing to the tyrant’s interest in superiority, or desire for
victory, Simonides brings about the tyrant’s spontaneous and almost
joyful recognition of all the shortcomings of his life and therewith a
situation in which the offering of advice is the act, not of an awkward
schoolmaster, but of a humane poet. And besides, in the moment that
Hiero becomes aware of his having walked straight into the trap which
Simonides had so ingeniously and so charmingly set for him, he will be
more convinced that ever of Simonides’ wisdom.

Appendix A

“The baby cried”

[Notes for lecture 1 ]

The materials to be discussed: The story fragment “The baby cried. The
mommy picked it up.” CTS p. 35 1

Plan

1 I will first make some observations about the segment

2 Then I will embark on the building of an apparatus for its analysis:

a That will first involve a consideration of what the apparatus ought
to do.

b Then the apparatus itself

3 Upshot of the constructed apparatus

4 A consideration of the possible fact that the segment is from a ‘story:’
How is it a ‘story;’ criteria for identification

5 Some further general considerations.

1 Observations

a While this is a piece of talk, we would in the first instance feel assured
in saying that there are two sentences here: SI, S2. Let us note that there are
also two actions, Al, A2, and we hear the story as providing that SI reports
Al, and S2 reports A2. Thus we hear A2 as following Al. Now this cannot
be accounted for simply by the fact that SI precedes S2:

a. 1 For, for one, SI is composed of three words, and we do not hear the
things those words name as following each other in the same way, though w3
follows w2, which follows wl.

1 CTS is the book Children Tell Stories.

These are the original Sacks materials, with a few changes. For example, the book
Children Tell Stories is sometimes cited - as CTS - but sometimes only the page
number is given. There, the notation CTS is added. Or, for example, whereas
elsewhere corrected errors have been preserved (e.g., Spring 1971, May 21), here
only the corrected version is shown. For example, in “Let us note that if there are also
two actions, Al, A2, we hear the story as . . .” the “if” is crossed out and an “and”
is inserted after “A2.” In the edited version it is shown as “Let us note that there are
also two actions, Al, A2, and we hear the story as . . .” And references to the
literature have been revised.

223

224

Part II

a. 2 Also, it is perfectly obvious that we can find elsewhere two

sentences linked as these are, with nothing between, where we would not
hear such an action sequence. For example, “The piggie got hit by the
choo-choo. He got a little hurt. He broke his neck. He broke his chin.”
CTS p .30.

a. 3 Furthermore a connective which particularly does the relating of
reported occurrences, ‘then’ - “This Indian caught this Indian. Then he
tumbled” - is not used, and seemingly is not missed.

Thus, it is indeed a problem how the relation between si, s2 as one of the
latter following the former provides information about the relation of al, a2.
How is it done?

When we develop the apparatus for analyzing the sentences some help may
be gotten on the question for the particular (sorts of) reported occurrences;
here some fairly general considerations.

We said initially there were two sentences. The determination that there are
two sentences must be made by reference to this fragment as talk, and not via
a consideration of the fact that the grammatical marker, a period, stands
between them. The child did not dictate, but told, and the child probably
doesn’t read or write, so the use of the period is not a resource for it. The
receiver of the stories might for one be wrong in putting the period where she
did, and we shall see that at other points there is good reason to suppose that
she is wrong in her transcriptions. But these facts are somewhat secondary to
the question of what is it that would give us grounds for saying that
something is or is not a sentence. Let us consider the relevance of our
observation (a); to wit, that we hear what we have called si, as a unit in the
sense that we do not take it as relevant that it comes out in a sequence of
words, i.e., we hear it as a unit and not as a left-to-right string. We hear s2
in the same way; and this is most clear since we do hear the relation between
s 1 , s2 as one of left to right.

If the way the piece is heard is related to the way it is generated, that si
is generated in some way, then the speaker begins again and generates s2,
where the sequential relationship is of relevance between but not within the
units. Other considerations are relevant within the units. That, or something
like that, could be a criterion for saying there are two units - sentences. We
leave aside for the moment the possible importance of the possible fact that
there is some strong relation between the production unit and the detection
unit. On the latter I refer you to Hockett’s paper, ‘Grammar for the
hearer’, in Roman Jakobson (ed.), Structure of Language and its Mathemat-
ical Aspects, vol. 12, Proceedings of Symposia in Applied Mathematics
(Providence: American Mathematical Society, 1961), 220-36. We will
come back to this sort of convergence problem, though probably not this
particular one.

If there are heard, and are generated, two internally structured units, then
the possibility of SI being relevantly followed by S2 would be provided for,
though the relevance of that, i.e., as providing a resource for indicating and

225

Appendix A

detecting the relation between the actions reported in each, would (not) 2 of
course be necessary. Note: ‘then’ could serve to do the job if a single sentence
were used; and’ alone might be quite ambiguous.

a Caveat or a query on what implication there seems to be, if we hear a
sequence of A 1 , A2 provided for in part by S 1 , S2 about the temporality of
the sequence - how soon after?

b Both the word ‘baby’ and the word ‘mommy’ are adequate ways of
referring to persons. A single category can be referentially adequate (economy
rule). 3

c Quite apart from the particular actions reported, ‘baby’ and ‘mommy’
are related categories. They are both categories from the same collection of
categories, a collection consisting of such as (mommy, daddy, baby, etcetera).
This collection consists of a group of categories where:

One important way the group or collection of categories is employed is to
provide a unit which, for the domain of relevance of the collection, locates a
set of persons organized by reference to a case of the unit, where the unit is
duplicatively operative across the community of culture users, such that the
possessive pronouns locate a set of persons ensconsed in particular cases.

Not every collection of categories is so organized; teams are, but of the
collection race; for that one, if there is some organization of Members,
persons, then it is regularly vis-a-vis incumbency in the same category.

The character of the collection’s organization is most simply evident in the
fact that when we hear the two sentences we take it that (‘we,’ i.e., members
of the community of culture users), when Members hear, they take it that it
that it is not some baby and some mommy who are involved, but the baby
of the mommy, and the mommy of the baby. Such a relation is not heard in
such a sequence as: “The next one who wakes up is the mailman. The next
one who wakes up is the doctor. “CTS p. 71. Note however, re the passing
reference to teams: “They rode on a choo-choo train. They went to a railroad
station. The conductor jumped out and said, “All aboard.’ And he went in
and the conductor was Mr Conductor and the engineer was Mr Engineer and
the fireman was Mr Fireman.’’ CTS p. 42, cf. further. Cf. p. 71 the three
mentioned after ‘doctor’ are not initially ensconsed in a single train, but it may
well be that they are heard as related.

c.l The consistency rule: Note an immediate use of it. ‘baby’ is not only
a category of the collection noted above; it is a category of the collection ‘Stage
of life’ (baby, child, adult, etc.). Given the two sentences it is, however, heard
as a member of the collection ‘family.’

Perhaps: If a category is a member of alternative collections, it is heard,
when used, as a member of that collection which the consistency rule locates
- thus the ambiguity problem may be resolvable if at least two categories are
used. Cf. p. 71, ‘fireman’ is an ambiguous occupational term - there is a guy
on trains, and a guy who deals with fires. The reference is clear by reference

2 The original is unclear.

3 The words ‘(economy rule)’ are handwritten.

226

Part II

to the preceding categories being train personnel. This is even more apparent
in the quotation, p. 4, from p. 45. 4

c.2 Note then the character of the consistency rule; on the one hand it is
used to generate the terms, given a first, and on the other it is used to detect
the relation of the generated terms. Cf. remarks in a. above. These parallels
shall turn out to be deeply important. A culture is an apparatus for generating
recognizable actions; if the same procedures are used for generating as for
detecting, that is perhaps as simple a solution to the problem of recogniz-
ability as is formulatable.

d For the given actions reported, the particular pair of person references
used are peculiarly relevant; that is, even sticking within the collection from
which they are chosen, they are particularly relevant. I will consider this fact
in the discussion of the apparatus.

e The use of ‘it’ for reference to the baby. Why is it that ‘it’ is used; note
that it is used elsewhere in the stories when that reference - to ‘baby’ - is
made. Relevance of boy - ‘he,’ girl - ‘she,’ baby -?

Also, and much more importantly, the occurrence of the pronoun ‘it’
means that at least to some extent the collection - pronouns - is available to
the kid. That means right off that two collections are held; they are, we shall
propose, quite different sorts of collections.

In modification of the discussion in ‘The search for help,’ I want to propose
that sticking to pn-adequate devices, we ought to divide them into two
classes; class two consists of (a) pronouns, and (b) proper names, and class one
of the rest. Now if it is the case that insofar as the consistency rule provides
that reference once made to a category from a collection, reference to others
should be made to them via categories of the same collection and not other
collections, then that rule only holds as between categories of class one. That
is, class two categories may be used in combination with class one categories.
The case at hand evidences this, and many others.

The following quote is from Weir, pp. 73-4. RH

Pronouns. Analyses of standard English pronouns vary in their interpre-
tations of the various pronominal forms as variants of a single
morpheme, or as separate morphemes. This problem is not at issue here
since we are primarily interested in the forms which occur, and in how
well they are established. Of the personal pronouns, the form it has by
far the widest occurrence (over one hundred) and is firmly established in
the child’s speech pattern. Occasionally it is over-used as a substitute for
a plural noun at the expense of them. The frequent use of it confirms the
child’s classification of nouns into personalized and non-personalized
classes as was pointed out. The non-personalized nouns permit substi-
tution by the pronoun it rather frequently, whereas the personalized
nouns (including names of persons, favorite toys, and animals) rarely are
substituted for by pronouns.

4 On the original p. 4, a quotation is shown from ‘p. 42.’ (See p. 225.)

227

Appendix A

In speaking of himself, the child uses Anthony much more frequently
than I, the use of which is restricted to special situations. The instances
where I does occur are either instances of complete quotations from
adult speech, like I don’t hear you where the use of do in a negative
indicative sentence also points to a non-internalized structural pattern; or
/ occurs in the child’s attempts at adult speech as apparent from the
context surrounding the sentence, e.g., Door. I’m fixing the door. Door’s
open-, or I appears when the rest of the sentence contains no particularly
bothersome structural consideration for the child, as in I have a blanket
like a lipstick where blanket like a lipstick is one of the child’s ‘set’
expressions as it were. However, the pronoun is also consciously
practiced as the two consecutive sentences show; I’m taking the yellow
blanket, I gave the yellow blanket.

Of about the same order of frequency is you, also about twenty-five
times, and its usage shows similarily to that of I. They are also practiced
together, for example, in this question-answer interchange: Where are
you going. I am going. The other personal pronouns occurring are she, her,
he, him, me, we, and them, all with a frequency ranging from one to
fewer than ten occurrences. All these pronouns appear in the correct
syntactic slots as compared to standard English, but the category of
gender has not yet been learned, similar to the case of possessive
adjectives. There is one occurrence of himself in the corpus, and we have
no additional information on the reflexive paradigm except perhaps for
the sentence, I told me, conspicuous by the non-use of the reflexive. It is
certainly not surprising to find pronominal substitution with the child’s
class of personalized pronouns so poorly learned in that it involves
viewing the speech event not only from the point of view of the
addresser and addressee, but also from the point of view of the person
or thing being discussed.

f Having made an observation about ‘it,’ one about ‘the’ is apposite;
consider the use of ‘the mommy.’ Quine, most particularly in Word and
Object, sec. 19, proposes: ‘A singular term, e.g., “mama,” admits only the
singular grammatical form and no article.’ intendedly referring to ‘our
perfected adult usage’ (p. 90). Now if it is the case that Quine is correct for
adults, it is quite apparent - we can find many more occurrences than just the
piece at hand - that children behave differently. Is there some special
explanation for this? We would of course, in doing this, have to come to some
discussion of his notion of ‘singular terms’ and ‘general terms’ and a
consideration of the relevance of these formulations. But it may well be so
adults do not, and children do, use ‘the mommy.’ If adults don’t, then
perhaps children ‘construct’ the usage - cf. here, Brown’s papers in which the
‘construction’ phenomenon is examined - in Lenneberg, and in Brown and
Bellugi.

Possible account; for adults there is a large selection, or a fairly large
selection, of devices which provide categories for handling any given person;

228

Part II

this permits perhaps the specialization of certain terms, such that for any given
person that term may only get used for some proper person(s). Now children
do not have such an armory. The basic device they do use is that ‘family’ one;
they make a big leap with the addition of the pronouns. Now they use the
family to do rather more extensive work than others do, and adults assist
them in this - referring for example to a large bunch of men and women
encountered as ‘uncle ’/‘aunt.’ The duplicative character of the family-unit set
of categories is quite important for the kids; for one, in combination with the
possessive pronouns it permits sorting out to go on while reference to a larger
bunch of people via the same unit is made. But it does involve their use of the
terms in becoming singular, as categorials; and ‘mommy’ gets such usages.
They say: “There are two mommies and a baby” and other like things, and
they do say ‘the mommy.’ The expansion for men (if ‘mommy’ is the normal
expansion of women who are not my mommy or aunt - they are then ‘a
mommy’ or ‘X’s mommy’) involves the use of the set of occupations to
handle ‘other men’ than daddy and the uncles - the milkman, etc., are ‘men.’

It might not then be too misleading to put it that the singular term
reference status of ‘mama’ is an adult luxury.

It might not be amiss either to notice how much is indeed handleable via
the family unit.

We shall of course consider the great leap forward gotten when the
pronouns are acquired. That anon.

g In the case at hand the fact that the characters are introduced via the
consistency rule’s use may be claimed to be a rather trivial consequence of the
possible fact that the child possesses the use of only a single class one
pn-adequate device. If that were so, then of course the consistency rule would
be used.

g. 1 Now, however, if the child learns something which for the single
device is trivially so, but which is not obviously so, and surely not trivially so
when more devices are added to its repertoire, but is nonetheless proper then,
too, then that is something not so trivial. So it shall be found.

g. 2 Now it is also non-trivial that in using the single device, it is used
duplicatively. The child is first given what turns out to be a case of a
duplicative unit; and it comes to see and use the categories as duplicative; for
this child in the next story we find the same device being used for ‘an
elephant’ though in that one we do get an introduction of something apart
from the device, i.e., ‘milkman.’ The duplication is not obviously the child
could come to see, as is done elsewhere, the whole relevant world as one
family.

h. Rather more generally we want to observe what is among the most
obvious of features of the fragment; it is recognizably what we may call ‘a
potentially correct description.’ The question of what is involved in a
potentially correct description, or in this being a potentially correct descrip-
tion, will be considered in the consideration of the apparatus. Here such
matters are to be noted, as: Adequate means of referring to characters,
adequate means of referring to characters’ actions, adequate pairing of

229

Appendix A

character references and actions, adequate pairing of the sentences, adequate
pairing of character reference terms and pairing of action terms, and of
combined pairings.

2 The apparatus

a What it should do

1 It should generate, interestingly, this segment.

2 It should permit us to discriminate this segment as a potential description
of a proper sequence of occurrences, from some segment which is an
improper sequence of occurrences properly described.

3 It should exclude the production of grammatical ‘non-potential descrip-
tion.’ In this point is something of the largest importance. We are going
to aim at building an apparatus which involves building constraints on
what an adequate grammar will do, such that what an adequate grammar
will do, some of the things it will do, we are going to rule out, and provide
for the non-occurrence of.

Our apparatus ought to give us this segment as a potential correct
description and as a potential correct description of a proper sequence. We
shall want to get potential correct descriptions of improper sequences, by a
suitable modification; and we shall want to see whether we need provide for
potential incorrect descriptions.

Appendix B

“The baby cried"

{ Notes for lecture 2]

Because of some problems with the room plug, lecture two did not get
recorded. Furthermore it was not laid out in the notes. In this I will very
briefly outline what was done; the next go round it will be recorded.

It was basically concerned with the character of “the baby cried” as a good
beginning.

(1) Our first concern was to provide that, if children’s rights to talk are
restricted, then it is expectable that they will have some special ways of getting
to talk, and these ways may involve them in having a proper way of
beginning which provides them with rights to continue, and for the
correctness of their having begun in the first place.

The hypothesis that they do have restricted rights to talk was established
by considering the character of what, for the young ones, say three or so, use
as their - not merely prototypical but at some times almost universally used
beginning; the line ‘You know what?’ and equivalents. An analysis was made
of the elegance of this line.

(2) With the establishment of the hypothesis that the analysis provided, we
moved to consider other ways that they might establish a right to talk; the
occurrence of troubles - where, by that we meant occurrences such that if they
happen then an adult has some obligation to act, where then the occurrence
of talk was directed to indicating that the adult obligation ought to be
actualized.

(3) This issue of troubles was then tied to the notion of topic, i.e., things that
could be said where the thing was something which was conversationally
generative; on this matter, too, ‘trouble’ was seen to be rich; comparison, for
example, of the amount of talk generatable if, to the question ‘how are you’
the answer is ‘fine,’ with the answer ‘lousy.’

The relevance of a matter being a good topic was made by reference to the
issue of a story being ‘about’ something, and thence hopefully something
about which more than a little might be said, i.e., the item offered might be
generative. ‘Trouble,’ not merely a good thing to provide grounds for talk in
general in conversation, a first item, but also a first item which is generative
and therefore specially attractive for stories.

These are the original Sacks materials, with a few changes. See Appendix A, note
to p. 223 for details.

230

Appendix B 231

(4) This led to considerations like: It might be said, given the fact that the
children’s stories, especially the youngest ones, are so much about troubles
that that is what their fantasy lives contain. While this might be so, I want
to point out that it is not unique for children’s stories, but is a large part of
adult conversation, and more particularly it is the base focus of Western
literature. The complaint has been made about Western literature, as about
children’s stories, though the two may not have been tied together, nor the
richness of trouble as a topic and the poverty of normality as a topic much
observed; here Tolstoy’s observation: happy families are all alike; unhappy
families are unhappy each in their own way; that quote not exact - first line
of Anna Karenina. (‘Unhappy’ for Tolstoy readily seen to be equivalent to
acting improperly - see second sentences.)

(5) Extend remarks to notion that only by behaving immorally is one doing
anything; the answer to the question, ‘What are you doing.’ ‘Nothing,’ as
equivalent to something proper. Sowing wild oats as building stories for one’s
future.

In this sort of line of analysis, then, is shown the good beginning for
conversation, and then for stories, of “the baby cried;” also its generativity.

Part III

Spring 1966

Lectures on Conversation, Volume I, II Harvey Sacks
© 1995 The Estate of Harvey Sacks. ISBN: 978-1-557-86705-6

Note

The Spring 1966 lectures start off with two lectures which eventually became the
paper ‘On the analyzability of stories by children. ’ In J. J. Gumperz and D.
Hymes (eds), Directions in Sociolinguistics (New York; Holt, Rinehart and
Winston, 1972), 325—45, reprinted in R. Turner (ed.), Ethnomethodology
(Harmondsworth; Penguin, 1974).

Two versions of each of the two lectures are included in this volume. The first
of each is the unedited text of the original working transcript, prepared just after
each lecture was delivered. The second of each is the unedited text of Sacks’
revision. With these two versions of each of the two lectures, can be seen what
Sacks took to be a next operation on a lecture. It is infinitely removed from
anything an editor might do; not only far more formal, but often a matter of
further development of the work at hand.

The difference between the original and revised versions of the first two
lectures, not to mention the difference between those and the published version,
point up the fact that the unrevised lectures which make up the bulk of these
volumes are at best informal, often rudimentary — the raw beginnings of what
they had in all too few cases a chance to become.

235

Lecture 1

“The baby cried. The mommy picked

it up”

The data is the first couple of sentences from a story in the book Children Tell
Stories, told by a girl two years and nine months old. It’s on page 35. I’ll read
it out to you. It says, “The baby cried. The mommy picked it up.”

What I’ll be trying to do for the beginning week or so is to give you some
idea of the sorts of things I’ll be trying to do, in what will be much greater
detail, I hope, throughout the rest of the semester, once we get the materials
dittoed up that I can give you, and once you get some familiarity with just a
few of the ideas we’re going to use.

Now let’s begin off by just making some observations about those two
sentences. And if you feel that what I say right now is the rankest kind of
subjectivism, then just hold off your feeling that that’s so, to see whether it
won’t be possible for what I say to be defended and used.

When I hear “The baby cried. The mommy picked it up,” one of the
things I hear is that the mommy who picks the baby up is the mommy of the
baby. Now it’s not only the case that I hear it that way - and of course there’s
no genitive there to say ‘its mommy picked it up,’ ‘his mommy,’ ‘her
mommy’ - when I hear it that way a kind of interesting thing is that I also
feel pretty confident that all of you, at least the natives among you, hear that
also. Is it some kind of magic? One of my tasks is going to be to construct an
apparatus which will provide for that fact to occur. That is, how it is that we
come to hear it in that fashion. Now exactly what else that apparatus will do,
we’ll see.

There’s another thing. I suppose we would say that there are two sentences,
and of course the fact that there’s a period doesn’t guarantee that. The child
doesn’t know how to write, didn’t put in a period. The person who was told
this story wrote it down and wrote a period in. But I take it we hear two
sentences, and we also hear the following kinds of things. Call this SI, call this
S2. This reports an occurrence, Ol, this one reports another occurrence, 02.
Now we hear the following sorts of things. We hear that 02 follows Ol like
S2 follows SI. And not only that. We also hear that 02 occurs because of Ol.
That is, the explanation for 02 occurring is that Ol did. And what we want
out of our apparatus is that it show how we hear that, also.

What we want to do then, speaking kind of roughly right now, is to
provide for how it is that something that’s done is recognized for what it is,

Except for the title, this is the unedited text of the original transcript.

236 Lectures on Conversation, Volume I, II Harvey Sacks

© 1995 The Estate of Harvey Sacks. ISBN: 978-1-557-86705-6

Lecture 1

237

where the big thing is that some action is recognized as a something, as a case
of some activity. What the something is, we’ll give a name to in a bit. In this
case where we can at least suppose for the moment that for some actions,
anyway, that persons do, they’re done in such a way as to be recognized, so
that the apparatus that can provide for how they’re recognizable can also
constitute procedures for generating the occurrence as a recognizable action, or
set of them.

Now if I asked you to explain the set of observations I made - which any
of you could have made just as well, and therefore they’re not proposed as
findings of mine, but as we’ll see, they only pose the problems that we have
to come to terms with - I suppose you might say something like the
following. We hear that it’s the mommy of the baby because she’s the one
who ought to pick it up, and if she’s the one who ought to pick it up and it
was picked up by somebody who could be her, then it’s her. And
furthermore, of course, the baby ought to have been picked up, given that it’s
crying - or something like being picked up, soothed in some way. So what
was done is just what ought to have been done.

Now let’s notice something that is worthy of some thought. What we’ve
been essentially saying so far is that the pair of sentences seem to satisfy what
a member might require of some pair of sentences for them to be a
description. And they’re quite recognizable as at least possible descriptions
without, for the particular case, inspecting what they might describe. You can
tell that they describe something just by looking at them. They ‘sound like a
description.’ And something can sound like a description. So some form of
statement can have the character that it could be a description, and you don’t
have to look at the circumstances to say, in the first instance, that it could be
a description.

What does that mean? One thing it means is that if the apparatus we
construct is directed to producing recognizable descriptions, then we, as
analysts, needn’t come to terms in the first place with what it is that babies
and mommies do to each other, to decide that it’s a description. We’re dealing
with the fact that this thing is recognizable as such without one having to look
at some circumstances it purports to talk about, or might be said to purport
to talk about, and therefore there may be ways of generating ‘possible
descriptions’ such that they are recognizable apart from anybody’s inspection
of the circumstances they may talk about.

How could that be possible? Let me just add here, when I talk of
generating these events and building an apparatus that generates them, the
name of that apparatus is, of course, a culture. At least the way I’m going to
use it. Now you ought to get some sense right off — and the reason I go
through the observations in the way I did is to give you some sense - of the
fine power of it. It’s not just producing brains that are roughly alike, but
they’re alike in fine detail. After all, these sentences are not the most
important set of events in the world. It’s a kind of very minor little thing, and
yet the operation of this culture, whatever it will look like, is such that you
all hear just what I said you heard, and I don’t even know you. That means

238

Part III

we’re dealing with something real and powerful. And not just grossly
powerful, like it provides for the rate of industrial development, but it
provides for little tiny things that God might have overlooked, perhaps.

Now let’s then try to begin to get some idea about what it is we want to
do. I’m going to introduce a bunch of terms that we need. You needn’t write
them down very carefully. You needn’t write them down at all, perhaps,
because most of the terms I’ll introduce right now you’ll find in the first ten
pages of this paper I assigned you. 1

The first term is ‘membership categorization device.’ I’ll just talk about it
most of the time as ‘categorization device.’ That’s the basic thing I’m
interested in, ‘membership categorization device.’ And those things are
collections of categories for referring to persons, with some rules of applica-
tion. So there’s a collection of categories and there are rules of application,
where these devices can be applied to populations and members apply them
to populations to say things about them, like that that’s ‘a baby.’ Typical
collections are things like (sex: male/female), (race: white/Negro’), etc.

Now, not any set of categories are a collection. We only talk about a
collection when the categories that compose it are categories that members do
in fact use together or collect together, as ’male’ and ‘female’ go together. A
set could be any bunch of categories.

Now the first thing we can note about this is, ‘baby’ and ‘mommy’ are two
categories from one collection, the collection you could roughly call ‘family,’
which consists of ‘baby,’ ‘mommy,’ ‘daddy,’ and a variety of other things.
Now there’s some rules which I’m going to use a great deal of the time. Let
me say in passing, I want you to notice that the term ‘collection,’ having the
character that it is categories-that-go-together, we could propose that some
categories go together and be wrong. So it’s not simply a construct of my own
in the sense that if I’m going to call it a collection, I am going to say whatever
it is that goes in there. The terms are constructed so as to best come to terms
with what it is that people in fact do, how in fact they handle these things.

Now one very simple thing, though we have to note it, is that any single
category from any of these categorization devices is adequate to refer to a
person. By which I mean, for one thing, that if a person uses a single category
to refer to somebody, then you don’t wait for them to continue, when they
said ‘baby,’ to give you more before you could have some idea of what in the
world they’re talking about. It’s perfectly imaginable that you would need to
have a string of these things put together so as to get some idea of the fact that
it’s really a person. We could imagine a world where you had to pick one
category, say, from each of three collections, to characterize a person in such
a way that anybody could know what you were talking about. But as it
happens, one category is enough, and ‘baby’ and ‘mommy’ are enough.

Now one basic rule in the use of these things, I’ll call the ‘consistency rule.’

1 Probably a draft of ‘An initial investigation of the usability of conversational data
for doing sociology,’ now in D. N. Sudnow (ed.), Studies in Social Interaction (New
York: Macmillan, The Free Press, 1972), 31-74.

Lecture 1

239

It holds that if you are categorizing some population of persons - if a Member
is categorizing some population of persons - then if they’ve used one category
from some collection for the first person they’re going to categorize, they may
- it is legitimate, permissable - to categorize the rest of the population by the
use of the same or other categories of the same collection. Now that can be
done with some of the devices. Some of the devices are so constructed that
they can categorize any population, like ‘sex’ can categorize any population,
however large, with its two categories: Male, female, male, female, female,
male, female, etc., etc. Some collections can’t do that, and for them, if you
don’t know what kind of population you have, it may well be that you cannot
use any given device to categorize them. We’ll get at the properties of these
various devices as we go along.

So the consistency rule says right off that if the first term is ‘baby,’ then
further persons may be referred to by such other categories as ‘mommy,’
‘daddy,’ and the like, i.e., it will tell us that if a population is being
categorized by persons, then, for one, the categories that they use will not be
a random set, but if we have a first, we can lock in on those that they will in
fact use for handling the rest. And if you look through the stories you’ll find
that that’s so, and if you listen to people you’ll find that that’s so.

Now I said ‘may.’ That’s a weak version of the consistency rule. A strong
version, which sometimes holds, says ‘must.’ Where it holds, we’ll consider as
we get along. That is, it says if some person has been categorized by some
category of some collection, then the rest of that population has to be
categorized by the same collection of categories.

Now I want to show you how we can now already get some kind of
interesting things to see. ‘Baby’ happens to be an ambiguous term under one
way I’m going to formulate ‘ambiguous’ - and it’s not an ambiguous way to
formulate it. That is, ‘baby’ is a term that occurs in at least two different
collections. For one, it occurs in the collection which you could call ‘stage of
life’ if you like: ‘Baby,’ ‘child,’ ‘teenager,’ ‘adult,’ whatever categories are in
it, and also it occurs in this collection we call ‘family.’ Now the fact is that you
don’t hear it as ambiguous. You don’t hear it as being ‘a very young child
cried, the mommy picked it up.’ What you hear is ‘The baby of the mommy
cried, the mommy of the baby picked it up. ’ Why? Well, what happens is the
consistency rule is not simply a rule that a categorizer uses to categorize
populations. It also can apparently be used as a maxim for hearers. Don’t
forget, we’re dealing with recognizability, constantly. And as a maxim for
hearers, what it can do is to say the following. If there are two categories used,
which can be found to be part of the same collection, hear them as part of the
same collection - which is the way you hear them. Now you might, if you
like, recognizing that in fact ‘baby’ is ambiguous in just the way I’ve said,
construct an explanation for yourself of why it is that you don’t hear it as
‘baby’ from the ‘stage of life’ collection. This ambiguity, of course, can be a
matter that you can play with. So a mother will say, ‘You may be 40 years
old, but you’re still my baby.’ And it’s not a queer usage, to use ‘baby’ in that
way.

240

Part III

Okay. So the consistency rule will then give us, given the first term, the
second - or some set of alternatives of which the actual second is one - from
a very delimited set. Now let’s consider this particular collection, ‘family.’ It’s
one of the series of collections which you can think of by the more prototypical
name of them, which is ‘teams,’ which have as one of their central properties
that they are what I’m going to call ‘duplicatively organized,’ by which I
mean the following. When they’re used on a population, what you do is you
take the set of categories of the collection, treat the set of categories as a unit,
and put people into cases of that unit, where what you don’t get is seven
daddies, six babies, three mommies, etc., but you get sets of people who are
organized into more or less complete units. So you end up with whole
families, families without fathers, families without babies, etc., and not an
enumeration of categories of persons. And if you counted that population,
you would very likely count it in terms of the number of families, and not the
number of mommies, daddies, babies, etc.

Now, when you have a collection that has this duplicative organization to
it, then it seems to be the case that if we give a list of terms in which categories
are used from such a collection, whichever one it may be, and we give the
terms various predicates, i.e., those people are doing various things. The third
baseman got up and scratched himself. The second baseman sat down, etc.
The engineer looked out the window. The conductor took a ticket. For each
of them, by virtue of the duplicational organization and the way it operates
on you, you hear, if you can - that is, if there’s not some reason why you can’t,
you always hear them as belonging to the same case of that unit. That is to
say, in just the way that you hear ‘that baby of the mommy,’ if I had a string:
The second baseman, third baseman, etc., you hear that as being ‘the second
baseman on the same team as the third baseman.’ And if you look at some
of the stories in which the kids will go through a string of titles - like one of
them has train jobs - you hear them as the same train. That then gives us at
least in part how it is that we hear ‘the baby’ and ‘the mommy of the baby.’

Now notice that it gives us this in a way that does not now at this point
depend on our feelings, and that in giving us this it also gives us a whole
bunch of things which are not in any obvious way just the same sorts of
objects. That is to say, first baseman, second baseman, pitcher, do not have,
except in some interesting way, a familial relationship to ‘family,’ and we can
lay out quite objectively what it is that routinely stands as the criterion for
deciding that something is duplicatively organized. If it is, then we can say
how it is that it’s going to be heard, and then of course, how it is that one may
speak so as to have it heard that way. For some things, for example, if I say
“The doctor got up, the engineer got up,” then if I'm going to try to suggest
that that engineer who happened to get up was the engineer of the doctor, I
have to tell you by using a genitive, for example. So these sets of rules, then,
are operating not only to provide for how this thing is done, but they’re usable
in various ways to provide for how it’s recognized.

Now one of the things you have to see is that what we’re talking about
now, at least the results and the phenomena, are not phenomena that you

Lecture 1

241

have to take my word for. It’s something that you can see. It’s stuff that
occurs. When you hear somebody do one of these things, listen to how you
hear it. Now the question of whether the account is adequate is another kind
of question. That doesn’t yet give us how we hear this thing to be a
‘description.’ More is needed for that.

Let’s introduce a term, which I’m going to call ‘category-bound activities.’
What I mean by that is, there are a great many activities which Members take
it are done by some particular category of persons, or several categories of
persons, where the categories are categories from these membership catego-
rization devices. Now how we would tell that they’re ‘bound’ is something
that I will give you as many different ways as we can use to tell, as I can. But
let’s consider, let’s say, one. I want to propose, as it’s obvious enough to you,
you wouldn’t argue with the issue, that ‘cry’ is bound to ‘baby.’ But we could
still of course ask, well, how would we decide if that’s so?

Now I’m going to propose that it’s bound to the other sense of ‘baby;’ that
is to say, ‘baby’ in ‘stage of life,’ since in this case ‘baby’ in ‘stage of life’ is
just as well probably ‘baby’ for purposes of ‘family.’ Okay. Now the ‘stage of
life’ categories are positioned: ‘baby’ . . . ‘adult’, with whatever in between.
Let’s see if we can describe a procedure for denigrating somebody, or for
praising them, the operation of which consists of the use of the fact that some
activities are category-bound. It would seem to be the case that if an activity
is category-bound, and it’s bound to some category that is in a positioned
collection, that given simply the positions and not necessarily the categories
themselves - perhaps it could work for any device which has that feature of
positions - the way we construct the denigration of somebody is to propose
about them that they are doing something that is bound to some category that
has a lower position than theirs. And alternatively, the way to praise them is
to propose that they’re doing some activity which is bound to a higher
category than theirs. And alternatively, you can simply say about them
‘You’re behaving like a . . . ,’ whatever the alternative position is. So if an
adult cries he can be said to be ‘behaving like a baby’ as a way of denigrating
him; if a baby does something about which babies are expected to cry and he
doesn’t, then you say ‘You’re behaving like a big boy.’ And we can pull these
things out and think of such things as ‘nigger work,’ for example - whatever
positional situations you want - as simply formal means which are tied to
positions for paising or degrading - which I introduced only to show that that
would be a means of seeing that a particular activity was in fact bound to
some category.

Now notice, that very thing will show us also that these collections are real,
in that what will be contrasted will be two positions from what we have
proposed to be a given collection. If an adult cries, somebody is not going to
say “You’re acting like a janitor.” They’ll say, “You’re acting like a baby” -
or, if it’s a male, “ . . . like a woman.” And again, this can be quite formally
analyzed because all you’re going to have to do is assign letters and numbers
to the positions, and say could we not generate a way of degrading somebody
by picking some of the activities bound to one of the categories in this position

242

Part 111

and using them as ways of characterizing persons who are recognizably
members of another. Try it and see. Though of course you already know it in
other ways.

Okay. So ‘crying’ is bound to ‘baby.’ The simplest way you make a
recognizable description is to take some category and some activity that’s
bound to it, and put them together. That sounds absolutely trivial. If you take
an activity that’s not bound to a given category and assert that it happens, or
assert it, people are perfectly willing to argue, to say ‘No.’ ‘No,’ even in the
case that we are just dealing with, a story. It’s not believable. If you read
literary criticism, people say about novels, “The characters are unbelievable.”
The whole possibility of, at least 19th-century, early 20th century literature
turns on the fact that descriptions are recognizably correct - interesting,
exciting, but recognizable apart from having to look at whatever it is that is
being characterized - if anything is being characterized - and that by playing
with the sets of properties and their relationships, one can in fact construct
cogent remarks about something. The fact that it is that kind of formal
operation is what would lend credence to the fear that, of course, computers
could build novels.

Lecture 1 (R)

(first revised draft, February 20, 1966)

“The baby cried, The mommy
picked it up”

In the first three lectures I intend, first, to present and employ several of the
more basic concepts and techniques which I shall be employing throughout
the rest of this course. Since most of those I shall use at this point may also
be found in the paper which is appended to these lectures, 1 these first may be
seen as reintroducing and extending the results developed there.

Second, I shall focus on the activity ‘doing describing’ and the correlative
activity ‘recognizing a description,’ activities which Members may be said to
do, and which therefore are phenomena which sociologists and anthropolo-
gists must aim to be able to describe. It will initially be by reference to an
examination of instances of Members’ describings that my attempt to show
how sociologists might solve their own problem of constructing descriptions
will be developed.

Third, and relatedly, I shall consider the problem of ‘correctness’ in a
manner parallel to the format that will be used to deal with ‘describing.’ I
shall examine a rather abstract statement which Members make and which
may be said to be correct, and in formulating how sociologists might proceed
to develop its correctness, a sketch of what correct sociological description
might look like will be exhibited.

Fourth, the analyses will permit us to redevelop the critique of current
social science that was first developed in the appended paper, to indicate again
its scope and the constraints it imposes on a sociology or an anthropology that
is not to suffer from the equivocality that preceding researches have suffered
from.

Proceeding in the fashion I have proposed will require both some
exceptional delicacy on my own part and on that of the reader. I proceed in
this way not only because it satisfies my own aesthetic but also because it will
permit a clear focus on several central, and neglected issues which social
science must face, most particularly, the problem of Members’ knowledge
and the problem of relevance. Let us then begin.

The initial data is the first two sentences from a ‘story’ offered by a two year
and nine months old girl to the author of the book Children Tell Stones. It

1 Probably a draft of ‘an initial investigation of the usability of conversational data
for doing sociology,’ now in D. N. Sudnow (ed.). Studies in Social Interaction (New
York: Macmillan, The Free Press, 1972), 31-74.

243

244

Part III

goes: “The baby cried. The mommy picked it up.” I shall first make several
observations about these two sentences. Before doing so, however, let me
note: If these observations strike you as a ranker sort of subjectivism, then I
ask you merely to read through just this first lecture to see whether it is or not
the case that the observations are both relevant and defensible.

When I hear, “The baby cried. The mommy picked it up,” one thing I
hear is that the mommy who picks the baby up is the mommy of the baby.
That is a first observation. (You will, of course, notice that the second sentence
lacks a genitive. It does not read, “Its mommy picked it up,” or variants
thereof.)

Now it is not only that / hear that the mommy is the mommy of the baby,
but I feel rather confident that at least many of the natives among you hear
that also. That is a second observation.

Is it some kind of magic?

One of my tasks is going to be to construct an apparatus which will provide
for the foregoing facts to have occurred; an apparatus, that is, which will show
how it is that we come to hear the fragment as we do.

Some more: I take it we hear two sentences. Call the first SI, and the
second S2; the first reports an occurrence, Ol, and the second reports an
occurrence, 02. Now, I take it we hear that as S2 follows SI, so 02 followed
Ol. That is a third observation. And also: We hear that 02 occurs because
of Ol; that is, the explanation for 02 occurring is that Ol did. That is a
fourth observation.

1 want the apparatus to show how we come to hear those facts also.

If I asked you to explain the group of observations which I have made,
observations which you could have made just as well - and let me note, they
are not proposed as sociological findings, but rather do they pose some of the
problems which sociology shall have to resolve - you might well say
something like the following: We hear that it is the mommy of the baby who
picks the baby up, because she’s the one who ought to pick it up, and (you
might eventually add) if she’s the one who ought to pick it up, and it was
picked up by somebody who could be her, then it was her, or was probably
her.

You might go on: While it is quite clear that not any two consecutive
sentences, nor even any consecutive sentences that report occurrences are
heard, and properly heard, as reporting that the occurrences have occurred in
the order which the sentences have, if the occurrences ought to occur in that
order, and if there is no information to the contrary (such as a phrase at the
beginning of the second, like ‘before that, however’) then the order of the
sentences indicates the order of the occurrences. And these two sentences do
present the order of the occurrences they report in the proper order for such
occurrences. If the baby cried it ought to have cried before the mother picked
it up, and not after. Hearing it that way the second is explained by the first;
hearing them as consecutive 2 or with the second preceding the first, some

2 This ought to be ‘simultaneous’ oughtn’t it?

Lecture 1 (R) 245

further explanation is needed and none being present we may suppose that it
is not needed.

Now let me make a fifth observation: All of the foregoing can be done by
many or perhaps any of us without knowing what baby or what mommy it
is that might be being talked of.

With this fifth observation it may now be noticed that what we’ve
essentially been saying so far is that the pair of sentences seem to satisfy what
a Member might require of some pair of sentences for them to be recognizable
as 'a possible description.’ They ‘sound like a description’ and some form of
words can, apparently, sound like a description. To recognize that some form
of words is a possible description does not require that one must first inspect
the circumstances it may be characterizing.

That ‘possible descriptions’ are recognizable as such, is quite an important
fact, for Members, and for sociologists.

The reader ought to be able to think out some of its import for Members;
for example, the economies it affords them. It is the latter clause, ‘and for
sociologists,’ that I wish to now attend.

Were it not so both that Members have an activity they do, ‘describing,’
and that at least some cases of that activity produced, for them, forms of
words recognizable as at least possible descriptions without having to do an
inspection of the circumstances they might characterize, then it might well be
that sociology would necessarily be the last of the sciences to be doable. For,
unless sociologists could study such things as these ‘recognizable descriptions,’
we might only be able to investigate such activities of Members as in one or
another way turned on ‘their knowledge of the world’ when sociologists
could employ some established, presumptively correct scientific characteriza-
tions of the phenomena Members were presumably dealing with, knowing
about.

If, however, Members have a phenomenon, ‘possible descriptions’ which
are recognizable per se, then the sociologist need not in the first instance know
how it is that babies and mommies do behave to examine the composition
of such possible descriptions as Members produce and recognize. Sociology
and anthropology need not await developments in botany or genetics or
analyses of the light spectra to gain a secure position from which Members’
knowledge, and the activities for which it is relevant, might be investigated.

What the sociologist ought to seek to build is an apparatus which will
provide for how it is that any activities, which Members do in such a way as
to be recognizable as such to Members, are done, and done recognizably. Such
an apparatus may be called ‘a culture.’ It will of course generate and provide
for the recognizability of more than just ‘possible descriptions’ and in later
lectures we shall be engaged in providing for such activities as ‘inviting,’
‘warning’ and so forth, as the data we consider permits and requires.

My reason for having gone through the observations I have so far made
was to give you some sense, right off, of the fine power of a culture. It does
not merely fill brains in roughly the same way, it fills them so that they are
alike in fine detail. The sentences we are considering are after all rather minor,

246

Part III

and yet the operation of the culture, whatever it will look like, is such that you
all, or many of you, hear just what I said you heard, and many of us are quite
unacquainted with each other. I am, then, dealing with something real and
something finely powerful.

We may begin to work at the construction of the apparatus. I’m going to
introduce several of the terms we need. The first term is 'membership
categorization device’ (or just ‘categorization device’). By this term I shall
intend: Any collection of membership categories, containing at least a
category, which may be applied to some population containing at least a
Member, so as to provide, by the use of some rules of applications, for the
pairing of at least a population Member and a categorization device member.
A device is then a collection plus rules of application.

An instance of a categorization device is the one called ‘sex:’ its collection is
the two categories (male, female). It is important to observe that a collection
consists of categories ‘that go together:’ For now, that may merely be seen as a
constraint of the following sort; I could say that some set of categories were a
collection, and be wrong. I shall present some rules of application very shortly.

Before doing that, however, let me observe that ‘baby’ and ‘mommy’ can
be seen to be categories from one collection: The collection whose device is
called ‘family’ and which consists of such categories as (baby, mommy,
daddy, etc.) where by ‘etcetera’ we mean that there are others, but not any
others, not, for example, ‘shortstop’.

Let me introduce a few rules of application. It may be observed that if a
Member uses a single category from any membership categorization device
then they can be recognized to be doing adequate reference to a person. We
may put the observation in a negative form: It is not necessary that some
multiple of categories from categorization devices be employed for recogni-
tion, that a person is being referred to, to be made; a single category will do.
(We do not by this mean that more cannot be used, only that for reference to
persons to be recognized more need not be used.) With that observation we
can formulate a ‘reference satisfactoriness’ rule, which we call the ‘economy
rule.’ It holds: A single category from any membership categorization device
can be referentially adequate.

A second rule I call the ‘consistency rule.’ It holds; If some population of
persons is being categorized and if a category from some device’s collection
has been used to categorize a first Member of the population, then that
category or other categories of the same collection may be used to categorize
further Members of the population. The former rule was a ‘reference
satisfactoriness’ rule; this latter one is a ‘relevance’ rule. We shall very
carefully consider the import of relevance rules in the ensuing lectures.

The economy rule having provided for the adequate reference of ‘baby,’ the
consistency rule tells us that if the first person has been categorized as ‘baby’
then further persons may be referred to by other categories of a collection of
which it is a member, and thus that such other categories as ‘mommy’ and
‘daddy’ are relevant given the use of ‘baby.’

While in its currently weak form and alone, the consistency rule may

Lecture 1 (R)

247

exclude no category of any device, even in this weak form (the ‘may’ form -
we shall eventually introduce a ‘must’ form) a corollary of it will prove to be
useful. The corollary is a hearer’s maxim. It holds: If two or more categories
are used to categorize two or more Members to some population, and those
categories can be heard as categories from the same collection, hear them that
way. Let us call the foregoing ‘the consistency rule corollary.’ It has the
following sort of usefulness. Various membership categorization device
categories can be said to be ambiguous. That is, the same categorial word is
a term occurring in several distinct devices, and can in each have quite a
different reference; they may or may not be combinably usable about a
single person. So, for example, ‘baby’ occurs in the device ‘family’ and also
in the device ‘stage of life’ whose categories are such as (baby, child,
. . .adult). A hearer who can use the consistency rule corollary will regularly
not even notice that there might be an ambiguity in the use of some category
among a group which it can be used to hear as produced via the consistency
rule.

It is of course clear that the two categories ‘baby’ are sometimes
combinably referential and sometimes not. A woman may refer to someone as
‘my baby’ with no suggestion that she is using the category that occurs in the
stage of life device; her baby may well be a full-fledged adult. In the case at
hand that problem does not occur, and we shall be able to provide the bases
for its not occurring, i.e., the bases for the legitimacy of hearing the single
term ‘baby’ as referring to a person located by reference both to the device
‘family’ and to the device ‘stage of life.’

With this, let us modify the observations of paragraph 2 of page 7 3 , as
follows: The consistency rule tells us that if a first person has been categorized
as ‘baby’ the further persons may be referred to by categories from either the
device family’ or from the device ‘stage of life.’ However, if a hearer has a
second category which can be heard as consistent with one locus of a first then
the first is to be heared as at least consistent with the second.

Given the foregoing we may proceed to show how the combined reference
of ‘baby’ is heard for our two sentences, and also how ‘the mommy’ is heard
as ‘the mommy of the baby.’ We shall deal with the latter task first, and we
assume from now on that the consistency rule corollary has yielded at least
that ‘baby’ and ‘mommy’ are heard as from the device ‘family.’ We assume
that without prejudice to the further fact that ‘baby’ is also heard as ‘baby’
from the device ‘stage of life.’

The device ‘family’ is one of a series which you may think of by a
prototypical name, ‘team.’ One central property of such devices is that they
are, what I am going to call, ‘duplicatively organized.’ I mean with that term
to point out the following: When such a device is used on a population what
is done is to take its categories, treat the set of categories as defining a unit,
and place Members of the population into cases of the unit. If a population
so treated is counted, one counts not numbers of daddies, numbers of

3 Coincides with paragraph 6, p. 246 above.

248

Part III

mommies and numbers of babies, but numbers of families - numbers of
‘whole families,’ numbers of ‘families without fathers,’ etc.

A population so treated is partitioned into cases of the unit, cases for which
what properly holds is that the various persons partitioned into any case are
‘co-incumbents’ of that case.

To these ways of dealing with populations categorized by way of devices
that are duplicatively organized there are correlative ‘hearers maxims.’ One
that is relevant to our current task holds: If some population has been
categorized by use of categories from some device whose collection has the
‘duplicative organization’ property, and a Member is presented with a
categorized population which can be heard as co-incumbents of a case of that
device’s unit, then hear it that way. (We shall consider the underscored phrase
shortly.) Now let it be noticed that this rule is of far more general scope than
we may seem to need, and in focussing on a property like ‘duplicative
organization’ it permits a determination of an expectation (of sociologists) as
to how some categorized population will be heard independently of a
determination of how it is heard. It is then formal and predictive, as well of
course as quite general.

Now, by the phrase ‘can be heard’ we mean to rule out predictions of the
following sort. Some duplicatively organized devices have proper numbers of
incumbents for certain categories of any unit. (At any given time a
nation-state may have but one president, a family but one father, a baseball
team but one shortstop on the field, etc.) If more incumbents of a category are
proposed as present in the population than a unit’s case can properly take then
the ‘can be heard’ constraint is not satisfied, and a prediction would not be
made.

The foregoing analysis shows us then how it is that we come to hear, given
the fact that the device ‘family’ is duplicatively organized, and the ‘can be
heard’ constraint being satisfied, ‘the mommy’ to be ‘the mommy of the
baby.’ It does of course much more than that. It permits us to predict, and to
understand how we can predict, that a statement such as: “The first baseman
looked around. The third baseman scratched himself. . .’’ will be heard as
saying “the first baseman of the team on which the third baseman is also a
player. . .’’ and the converse.

Or, putting the claim more precisely, it shows us how, in part. In part,
because for the materials at hand it happens that there are other means for
providing that the same hearing be made, means which can operate in
combination with the foregoing - otherwise sufficient ones - to further assure
the hearings we have observed. This will be done in the next lecture. Let us
now undertake our second task, to show how ‘the baby’ is heard in its
combined form, i.e., as the category with that name from both the ‘stage of
life’ device, and from the ‘family’ device.

Let me introduce a term which I am going to call ‘category-bound
activities.’ While I shall not now give an intendedly careful definition of the
term, what I shall do is to indicate what I mean to notice with it and then in
a while offer a procedure for determining for some of its proposed cases that

Lecture 1 (R)

2 49

they are indeed cases of it. By the term I intend to notice that many activities
are taken by Members to be done by some particular or several particular
categories of Members where the categories are categories from membership
categorization devices.

Let me notice then, as is obvious to you, that ‘cry’ is bound to ‘baby;’ that
is, to the category ‘baby’ which is a member of the collection from the ‘stage
of life’ device. Again, the fact that Members know that this is so only serves,
for the sociologist, to pose some problem. What we want is to construct some
means by reference to which a class, which proposedly contains at least the
activity-category ‘cry’ and presumably others, may have the inclusion of its
candidate-members assessed. We will not be claiming that the procedure is
definitive as to exclusion of a candidate-member, but we will claim that it is
definitive as to inclusion of a candidate-member.

It may be observed that the members of the ‘stage of life’ collection are
‘positioned’ (baby. . .adolescent. . .adult. . .), an observation which, for now,
we shall leave unexamined. I want to describe a procedure for praising or
degrading Members, the operation of which consists of the use of the fact that
some activities are ‘category-bound.’ If there are such procedures, that will
provide one strong sense of the notion ‘category-bound activities’ and also will
provide for any given candidate activity a means of warrantably deciding that
it is a member of the class of category-bound activities.

For some positioned-category devices it can be said as between any two
categories of such a device that A is either higher or lower than B, and if A
is higher than B, and B is higher than C, then A is higher than C.

We have some activity which is a candidate-member of the class
‘category-bound activities’ and which is proposedly bound to some category
C. Then, a member of either A or B who does that activity may be seen to be
degrading himself, and may be said to be ‘acting like a C.’ Alternatively if
some candidate-activity is proposedly bound to A, a member of C who does
it is subject to being said to be acting like an A where that assertion constitutes
‘praising.’

If we, using the ‘stage of life’ categories, subject ‘crying’ to such a test we
do find that its candidacy as a member of the class ‘category-bound activities’
is warrantable. In the case of ‘crying’ the results are even stronger. For, it
appears, if a ‘baby’ is subject to some circumstances which would for such a
one warrant crying, and he does not, then his ‘not crying’ is observable, and
may be used to propose that he is ‘acting like a big boy’ where that assertion
is taken to be ‘praise.’

The foregoing procedure can obviously enough be used for other devices
and other candidate activities. Other procedures may also be used; for
example, one way to decide that an activity is category-bound is to see
whether, the fact of membership being unknown, it can be ‘hinted at’ by
naming the activity as something one does.

S : So, you can’t watch television. Is there anything you can stay

interested in?

250

Part III

C: No, not really.

S : What interests did you have before?

C : I was a hair stylist at one time. I did some fashions now and then.

Things like that.

S : Then why aren’t you working?

C : Because I don’t want to, I guess. Maybe that’s why.

S : But do you find that you just can’t get yourself going?

C: No. Well, as far as the job goes?

5: Yes.

C: Well, I’ll tell you. I’m afraid. I’m afraid to go out and look for a job.

That’s what I’m afraid of . But more, I think I’m afraid of myself
because I don’t know. I’m just terribly mixed up.

S : You haven’t had any troubles with anyone close to you?

C: Close to me. Well, I’ve been married three times and I’m- Close, you

mean, as far as arguments or something like that?

5: Yes.

C : No, nobody real close. I’m just a very lonely person. I guess I’m very-

S : There’s nobody who loves you.

C : Well, I feel that somebody must some place, but I don’t know where

or who.

S : Have you been having some sexual problems?

C: All my life.

S : Uh huh. Yeah.

C: Naturally. You probably suspect - as far as the hair stylist and -

either go one way or the other. There is a straight or homosexual,
something like that. I’m telling you, my whole life is just completely
mixed up and turned over and it’s just smashed and smashed and I’m
not kidding.

Having constructed a procedure which can warrant the candidacy of some
activity as a member of the class category-bound activities, and which
warrants the membership of ‘cry’ and provides for its being bound to ‘baby,’
i.e., that category ‘baby’ which is a member of the ‘stage of life’ collection, we
move on to see how it is that ‘the baby’ in our sentence is heard in the
combined reference we have proposed.

We need, first, another hearer’s maxim: If a category-bound activity is
asserted to have been done by a member of some category where, if that
category is ambiguous (i.e., is a member of at least two different devices) but
where at least for one of those devices the asserted activity is category-bound
to the given category, then hear that at least the category from the device to
which it is bound is being asserted to hold.

The foregoing maxim will then provide for hearing “The baby cried” as
referring to at least ‘baby’ from the ‘stage of life’ device. The results obtained
from the use of the consistency rule corollary being independent of that, are
combinable with it. The consistency rule corollary gave us at least that ‘the

Lecture 1 (R) 251

baby’ was the category from the device ‘family.’ The combination gives us
both.

If our analysis seems altogether too complicated for the rather simple facts
we have been examining, then we invite the reader to consider that our
machinery has intendedly been ‘overbuilt.’ That is to say it may mm out that
the elaborateness of our analysis, or its apparent elaborateness, will disappear
when one begins to consider the amount of work that the very same
machinery can perform.

In the next lecture we shall attempt to show that the two sentences, “The
baby cried. The mommy picked it up,” constitute a possible description.

In the third lecture we will examine the central significance that the various
rules, such as the consistency rule and the hearer’s maxims, have for social
sciences. It is, we now but assert, by reference to these that the equivocality
which undercuts almost all previous social science research may be removed.

Lecture 2

“The baby cried. The mommy picked

it up”

We’re talking about these two sentences, “The baby cried. The mommy
picked it up.” I want to consider some of the import of the fact that some
activities are, as I put it, ‘category-bound.’ The import I’m going to consider,
relevant to the issue of a description, how they’re made, is somewhat separate
from coming to see some offered statement as a description or a possible
description, and concerns how, given some occurrence, one can come to
formulate what one takes to be a description of it, because that will in part
turn on the fact that some of these activities are category-bound.

Now an initial relevance can’t probably that easily be seen by reference to
this pair, ‘baby’ and ‘cried,’ but it’s present there as well, though it can be
more easily seen with respect to the next - ‘the mommy’ and ‘picked it up,’
and that concerns the issue of what I argue in the paper I’ve given you, and
what you can easily enough see when you consider it, that for any person
being characterized by a category from one of these categorization devices, it’s
always the case that alternative categories could categorize the same person,
where ‘alternative’ means, not from within the same collection, but by a
category from a different collection. So that if you use ‘sex:’ ‘male/female,’
you could say “the male cried.” Now you don’t. But let’s leave that aside for
the moment. Now the fact that activities, or some, are category-bound,
provides another relevance rule for choosing categories. The first relevance
rule I gave you was this consistency rule which said that if you had a category
assigned to a first person, then in categorizing second and third persons you
could select categories from the same collection that was used in the first
instance to categorize the first. That had a corellate - the hearing maxim I
gave you, i.e., that if there are two categories which can be heard as from the
same collection, hear them as from the same collection. Now, another
relevance rule turns on the following. It turns on this fact that some activities
are category-bound. Here’s one way it works. Let’s consider it, just for the
purposes of making the matter intuitively easier to see, by reference to this
second sentence. You’re standing somewhere and you see two people you
don’t know. Suppose one is a child. Suppose the second is a female adult. The
first cries. The second picks it up. What you see is that the one who picks it
up is the mommy of the baby.

Now again you could see in principle - that is, in principle you could apply

Except for the title, this is the unedited text of the original transcript.

252

Lecture 2

253

any number of categories to the same person - the fact that picking up crying
babies is in the first instance the business of the mommies of them operates to
have you see that as not merely ‘a’ mommy, but that mommy who is the one
who is supposed to pick up her baby. Now you can see her as the mommy
of that baby without in the first instance knowing that she’s either a mommy
or, in fact, the mommy of that baby.

Now what’s a consequence of that, or what are some consequences of that,
that are worthy of noticing? First of all, if it’s the case that what we would
conventionally call a ‘norm,’ what members of the society would call a norm
or rule of proper behavior, says something like mommies ought to soothe
their crying babies, then what seems to be present is not simply that that’s a
maxim of good behavior for mommies, but that one can use it to see who it
is that’s doing something which it can characterize. So these sets of norms,
then, can operate to provide for how it is that you observe some set of
activities, and see them not merely as correctly characterized by you, but as
some activity correctly occurring. By ‘correctly occurring’ I mean, you don’t
say “Who’s that lady who’s picking up that baby?,’’ where that could be
something very wrong. I’ll make that clear in a moment.

It’s a rather regular thing that happens which the foregoing can analyze for
us. You read it in the papers, it goes like this. A woman is assaulted in a car
on Wilshire Boulevard. She says, “All sorts of people passed by. Nobody did
a thing.’’ Some of the witnesses are gotten to, and they say “Yeah, sure I saw
it, but I figured it was her husband and they were having a family argument,”
which is something that’s okay for a husband and wife to have, and which
also provides that it’s none of their business to interfere. You see a man
standing on the street, say an adult male, and you see him smack a
15-year-old girl. That could be something that you ought to intervene in,
except, it would be fairly regular that in the first instance what you see is ‘a
father chastising his daughter’ and not ‘a strange man assaulting some strange
girl,’ where in either case you know neither of them. So what seems to happen
is that for activities that are category-bound what one sees when one sees a
pair, a triplicate, a single person, doing one of them, is that they are the person
who properly does that. Now that’s a different sense of ‘norms regulating
activities’ than one might usually use in doing sociology or anthropology,
where it would be said by and large that the norms are followed by those
persons who do the things - or ought to be. They provide also, then, for how
it is that persons who know neither of the people, come to choose what
categories they ought to use, or they do naturally use, to see what it is that’s
going on. And in that sense they provide rules of relevance for selecting
categories.

So again, now, if what we had is a scene, somebody who could be a baby
crying, somebody who could be a mommy, the mommy of the baby, picking
it up, then what you see is that taking place, and we get now another sense
in which this pair of sentences can be seen as a description in a way that can
also be seen to be ‘apart from identifying the circumstances;’ that is, apart
from investigation. One doesn’t ask, one doesn’t rely on having to know who

254

Lecture 2

those people are in the first instance. One can see it, and see it as correctly
characterized by you via ‘the description.’ Such a fact is of immense
importance because it would be the kind of violations of norms that would
not permit that to get done that would be, probably, much more consequen-
tial than those violations which would involve mommies not picking up their
babies. Were it the case that you had to decide in some way like going up and
asking these persons whether they were the proper set of people to be doing
whatever it is they’re doing, then the world might become much more
complicated than it is. And I will call, in principle, such things as people
‘passing’ under such rules, ‘subversion.’ When a woman walks away from a
supermarket with the baby carriage filled with a baby that’s not hers, that’s
the sort of thing I’m talking about with ‘subversion.’ It’s not seeable.

There are a variety of further things I want to deal with about these
sentences. One of them I’ll deal with at considerable length in a while, and it
concerns now not the characterized events, i.e., the things these may be
talking about, but what makes an adequate characterization of these - again,
independently of what it is that’s taking place - because that has some very
important consequences. “The baby cried. The mommy picked it up.’’ That’s
perfectly okay, and we don’t have to know in the first instance where the
mommy was when the baby cried, how it happened to come on the scene, or
anything else. The baby cried, the mommy picked it up. “The baby cried”
not only provides that the mommy ought to, but in some way provides that
mommy will be around to do it.

Now we’ll say that for some pairs of activities, pairs of actions that are
related by norms, that there’s at least a rule of adequate description which says
‘character appears on cue,’ i.e., if the first takes place and it’s an adequate
grounds for the second taking place, then it’s okay to describe the thing
without having provided for how it is that that second person happened to
come on the scene to do whatever it is they properly do, if one says the first
occurs and the second occurs as well. Now that seems like a very literary kind
of rule, and you might take it that it holds for the Greek dramas, maybe, but
not anywhere else, but that’s not true, and it’s also important that I just said
that there’s two actions related by norms. It does not merely involve a
situation where the second — the person who does the thing that’s second - is
bound to do it, given the first, but if the first is, for example a violation, then
you can provide that the second occurred. For example, “He was speeding
and he got arrested.” Perfectly okay. One needn’t provide for the fact that the
policeman somehow got on the scene.

Now I’ll try to squeeze out some consequences of that in a while, but it’s
a task, so let me just hold off. I only wanted to indicate to you that this kind
of thing - two occurrences properly tied together - the character-appears-on-
cue rule will make, then, that description adequate.

I’ve been talking alternatively of this as 'possible description’ or ‘possibly
correct description,’ and the technical term I’ll tend to use is ‘potentially
correct description’ for things like it. Now that’s going to be discriminated
from what we could call an ‘actually correct description’ which would be such

Lecture 2

255

a thing used on some occasion where it’s talking about something it
characterizes, in contrast to what it is here, perhaps, a story, or what it could
also perfectly well be, a lie. The power of good lies is that they could be good
descriptions, and what would differentiate the phenomenon ‘lies’ from ‘a
story’ would be that the one who tells the story makes no claim at all that he’s
talking about some actual occurrence. A liar is presumably engaged in passing
off his potentially correct description as actually correct. That suggests one
fairly neat thing. If you’re dealing with young children, one of the things
you’d like to know about them is, first, when do they have the competence
to lie? And what size of the things that they talk about do they have the
competence to lie about? Stories are, then, a very useful phenomenon; that is
to say, children’s stories are very useful for adults, since if a child can tell a
story which passes as a description, then you can figure that you better watch
out for those things that they do not propose, as such, as stories, but propose
as actual descriptions.

Now, in coming to see how this thing might be a story, it is of course
perfectly obvious, or it ought to be obvious, that the sheer fact that this lady
asked the kid to tell her a story doesn’t tell us that that’s a story. To suppose
that it does is to simply fall into the most characteristic error of social science,
which is only to interpret the answers to questions and not the questions. Now
what we have to do then is to come to some way of formulating what makes
something a story. Then, to see whether that formulation we make holds, i.e.,
characterizes the object we’re considering. It could be either that the child
doesn’t know what a story is, or it doesn’t feel like telling one. Maybe it
supposes that it has fool enough audience so that anything it says will be
considered its version of a story. Also, of course, we’d like to know does the
kid have a good idea of what it is that it takes to make a story? There are a
variety of reasons why that could be of interest. First, there are lots of parts
of stories, or some parts of stories, that are rather standardized cultural
objects, and we would like to know whether the kid has them under control
and knows that the way you build one is to use those things.

For example, ‘once upon a time.’ That’s a story-beginning. One thing
about some stories anyway is that they have special ways of beginning. That
is to say, one can differentiate between a ‘start,’ which could be anything, and
the fact that the first thing that occurs is a ‘beginning.’ And perhaps the last
thing that occurs is an ‘ending.’ Now if we start off with endings, then we find
that in this story, and in some of the others, it does look like the kids know
that there ought to be an ending. And again, I mean by ‘ending,’ something
different than, distinctive from, simply stopping, but that we can characterize
the thing as an ‘ending.’ If it is an ending, and it occurred in last position, then
we have an ending for the story. If it’s something that looks like a beginning
and it occurred in the first position, it looks like we have a beginning. We
could then say that the ending occurred too soon, as is often said, or that it
stopped but there was no end, it had no beginning, etc., etc., all those
possibilities where we have an object and a position.

Now what it takes to make an ending, I certainly am not going to

256

Lecture 2

formulate exhaustively. But if you look at, say, the close of this story and
others, then it looks like we’ve got an ending there, and it looks like we have
something that was built as an ending. The story ends, “and she went to
sleep,” i.e., the kid takes an end to something, not just anything either, a most
natural ending, i.e., an ending to the day, and uses it to end the story.

Now the problem of beginnings is a rather more complicated one, and
we’ll try to deal with it a little. The problem with beginnings for kids’ talk is
a special one because, we might say, first roughly and only as an assumption
to see whether it’s so - though naively it’s obvious - kids have restricted
rights to talk. That being the case, by assumption, let’s see whether the ways
that they go about beginning to talk can be most adequately seen to be
solutions to the problem of getting to talk under restricted rights. What we
want to do is see whether their beginnings are a more or less nice, methodical
solution to a problem which focusses on having to have a good way to begin.

Kids around the age of three — and that’s what we’re dealing with here -
go through a time when they have what may be almost a universal way of
beginning any piece of talk they make to adults. They use it just about every
time. It has a little bit of variation, but it’s a thing like “You know what,
Daddy?,” “You know something?,” things like that. What we want to do is
to analyze that, to see how it’s a beginning, and to see how, as a beginning,
its providing a solution to their having restricted right to talk informs us that
our notion that they have restricted right to talk is correct.

I’ll introduce, without data at this point because the facts are so simple and
obvious that you can check them out easily enough for yourself - you know
them anyway - a couple of rules of conversational sequencing, two-party
sequencing. Two-party conversations are a special basic phenomenon, and
what I say is not intended as applying to three-or-more-party conversations.
One basic rule of two-party conversations concerns a pair of objects, questions
and answers. It says, if one person asks a question, the other person answers
it. I will have to give characterizations of those objects, but I’m not going to
do it now. I’ll do it when we get to discussing conversation much more
seriously.

A second rule, and it’s an absolutely fundamental rule because by reference
to it the infinite character of a conversation can be found, is, the person who
has asked the question can talk again after the other answers, and can ask a
question. I call it the ‘chaining rule,’ and obviously it has as its consequence
that a conversation, in principle, its method, has no finite length but is just
extendable in this fashion: Q, A, Q, A, Q, A.

Now here’s a question: “You know what?” It has, as a question, an
extremely curious kind of answer, and the answer is one of those fairly rare
but not altogether rare objects which, as an answer, is also a question, i.e.,
“What.” With that as a question, the question-answer rule is turned around.
The kid’s been asked a question. It’s his business to answer. You can
characterize, then, the restriction as involving a right to use a first item, which,
however, in this case, is the first item that for anybody else is a chainable right;
that is, indefinitely extendable. Kids don’t use it that way. They use it simply

Lecture 2

257

to set up the other person in the position of questioner. They thereafter do not
proceed on their own say-so, or by reference to the chaining rule, but as
obliged to talk under the position of an answerer.

Now of course it happens to be that the question they elicit with their
question has altogether open form, and what stands as an answer is whatever
the kid takes it to be an answer, and thereby provides him with the
opportunity to say what it is that he wanted to say in the first place. However,
now not on his own say-so, but under obligation. That’s very nice.
Three-year-old kids are kind of clever. In that case, then, what it looks like is
that the kids take it they have a right to begin, to make a first statement, and
nothing much more. That being the case we can focus on their beginnings as
not just the first thing they happen to say, but as something which can be
examined for its status as a ‘beginning’ to see whether, again, it will be the
case that what it is that’s put in first position in the story is something that
stands as a beginning. “The baby cried’’ reports a piece of trouble. It’s a kind
of special item in itself, in that if we can talk of something like ‘once upon a
time’ as a token of a story-beginning, it just happens to be your beginning
because that’s the way stories begin. In that sense, a token. You give nickels,
whatever, because that’s what you give, as compared to this thing being a
‘beginning’ in some special way, by way of its properties.

Then we’ll introduce, as a nothing-special term, there’s another kind of
beginning, which we’ll call ‘tickets.’ Tickets are kind of interesting in that
they’re beginnings of the following sort.

Q: What was that word?

HS : Tickets.

Q: Tickets?

HS: Yeah. T-i-c-k-e-t. I use it for the following reasons. Suppose two
people, two adults, don’t have rights to talk to each other, in general.
They are not, under whatever formulation there may be - and I’ll try
to formulate such things eventually - persons who converse with
each other. There are conditions under which they can talk, or one
can talk to the other, and that those conditions are the conditions
being used to begin a piece of talk for them, is something they can
deliver in the first piece of talk, via a ticket. Which is to say that they
say something which, having been said, warrants that they’ve talked
in the first place. Like, “Your pants are on fire.” ((laughter)) It’s not
just an opening, but it’s telling you why I’m talking. The occurrence
of trouble is, then, that sort of object. Persons who have - now, not
simply children who have generally restricted rights, but for any pair
of persons who otherwise don’t talk to each other, they can use a
ticket as a beginning. One kind of ticket is the announcement of
some trouble. That holds for whoever it is you’re talking to, and
that’s, then, the sort of object that “the baby cried” is. If the baby
cried, then “the baby cried” is a perfectly good thing to say, anybody
can say it, including a child who otherwise shouldn’t talk, and

258

Lecture 2

should say it as a first statement. Okay, so we can find out how that’s
a beginning. If then it’s a beginning in general, it can be a beginning
of a story.

It has a little bit of side relevance that concerns the following. If
you look at all these stories, the overwhelming number of them are,
in one way or another, tragic, and we could read directly from them
into what we suppose the children’s fantasy lives look like and say
it’s filled with all sorts of horrors. At least a partial alternative,
though I certainly don’t claim to be saying what their fantasy lives
are full of, the character of the stories could be partially accounted for
by reference to those things that kids can properly talk of or have as
occasions for talking, which are, troubles. But that’s neither here nor
there.

I don’t have time now. I’ll continue next time. What I want to do
then is further spell out this business of ‘characters appear on cue.’
Now let me just say what I want to do, and I’ll say it because you
can pull features out of certain stories which will be in point. I’d like
to suggest the following. We may find from a set of stories that kids
may take it, young kids anyway, that the occurrence of the trouble
that involves them in crying or whatever, is sufficient to bring the
assistance, and that danger, then, is a rather different object than you
would like it to be for them. I pose a problem which is how do you
get them to see, not simply that the fact that they’re in trouble
doesn’t mean that they’re going to be pulled out, but that if they
want some people around, the simplest way is not to go jump in the
river or something like that.

Now if you look at a bunch of the stories and you take it that
they’re possibly mapable onto what it is that the kid figures is going
on, you may get some such idea, and you may try to work out some
way of shifting that around - if you don’t think there are already too
many kids around.

Okay. Continue next time.

Lecture 2 (R)

(first revised draft (February 24, 1966)

“The baby cried. The mommy
picked it up”

I am examining the two sentences, “The baby cried. The mommy picked it
up,” and in this second lecture I shall proceed in much the same manner as
I did in the first, beginning with some observations and then exploring their
import. I shall focus first on the fact that an activity can be category-bound
and then on the import of there being a norm which provides for some second
activity given the occurrence of a first, considering both of these with regard
to the ‘correctness,’ for Members, of ‘possible descriptions.’

Let me leave aside our two sentences and for the moment consider instead
some observations on how it is that I see, and take it you see, describable
occurrences.

Suppose you are standing somewhere, and you see a person you don’t
know. The person cries. Now, if I can, I will see that what has happened is
that a baby cried. And I take it that you will, if you can, see that too. That’s
a first pair of observations. Suppose again you are standing somewhere and
you see two people you don’t know. Suppose further, one cries, and the other
picks it up. Now, if I can, I will see that what has happened is that a baby
cried, and its mother picked it up. And I take it that you will, if you can, see
that too. That’s a second pair of observations.

Consider the first pair of observations. The modifying phrases, to deal with
them first, refer simply to the possibility that the category ‘baby’ might be
obviously inapplicable to the crier. By reference to the ‘stage of life’ collection
the crier may be seen to be an adult. And that being so, the ‘if . . . can’
constraint wouldn’t be satisfied. But there are certainly other possible correct
characterizations of the crying person. For example, without respect to the fact
that it was a baby, it could be either ‘male’ or ‘female’ and nonetheless I
wouldn’t, and I take it you wouldn’t, seeing the scene, see that ‘a male cried’
if we could see that ‘a baby cried.’

The pair of observations suggest the following ‘viewer’s maxim:’ If a
Member sees a category-bound activity being done, then, if one can see it being
done by a member of a category to which the activity is bound, see it that way.
The viewer’s maxim is another ‘relevance rule’ in that it proposes that for an
observer of a category-bound activity, the category to which the activity is
bound has a special relevance for formulating an identification of its doer.

Except for the title, this is the unedited text of the original revised draft.

259

260

Part 111

Consider the second pair of observations. As Members, you of course know
that there is a norm which might be written as: A mother ought to try to
soothe her crying baby. I, and you, not only know that there is such a norm,
but, as you may recall, we used it in doing our hearing of “The baby cried.
The mommy picked it up.” In addition to the fact of duplicative organiza-
tion, the norm was relevant in bringing us to hear that it was the mommy of
the baby who did the picking up. While we would not have heard that it was
the mommy of the baby for other pairs of activities in which the two were
involved (but not any pair) the fact that the pair were relatable via a norm
which assigns the mother of the baby that duty may have operated in
combination with the duplicative organization to assure our hearing that it
was she who did it.

Leaving aside the hearing of the sentence, we are led to construct another
‘viewer’s maxim:’ If one sees a pair of actions which can be related via the
operation of a norm that provides for the second given the first, where the
doers can be seen as members of the categories the norm provides as proper
for that pair of actions, then (a) see that the doers are such Members, and (b)
see the second as done in conformity with the norm.

This second viewer’s maxim suggests a rather fundamental observation
about norms. In the sociological and anthropological literature, the focus on
norms is on the conditions under which, and the extent to which, they govern
- or can be seen by social scientists to govern - the relevant actions of those
Members whose actions they ought to control. While such matters are of
course important, our viewer’s maxim suggests other importances of norms,
for Members.

Viewers use norms to provide some of the orderliness, and proper
orderliness of the activities they observe. Via some norm, two activities may
be made observable as a sequentially ordered pair. That is, viewers use norms
to explain both the occurrence of some activity given the occurrence of
another, and also its sequential position with regard to the other, e.g., that it
follows the other, or precedes it. That’s a first importance. Second, viewers use
norms to provide the relevant membership categories in terms of which they
formulate identifications of the doers of those activities for which the norms
are appropriate.

Now let me observe, viewers may use norms in each of the preceding ways,
and feel confident in their usage without engaging in such an investigation as
would serve to warrant the correctness of their usages. This last observation is
worth some further thought.

We may, at least initially, put the matter thusly: For viewers, the usability
of the viewer’s maxims serves to warrant the correctness of their observations.
And that is then to say: The usability of the viewer’s maxims provides for the
recognizability of the correctness of the observations done via those maxims.
And that is then to say: ‘Correct observations’ or, at least, ‘possible correct
observations,’ are ‘recognizable.’

(Members feel no need in warranting their observation, in recognizing its
correctness, to do such a thing as ask the woman whether she is the mother

Lecture 2 (R)

261

of the baby, or ask her whether she picked it up because it was crying. They
feel no such need - so long, that is, as the viewer’s maxims are usable.) In
short: ‘Correctness’ is recognizable, and there are some exceedingly nice ties
between recognizably correct description and recognizably correct observa-
tions.

One such tie which is relevant to the tasks we have undertaken is: A string
of sentences which may be heard, via the hearer’s maxims, as having been
produced by use of the viewer’s maxims, will be heard as a ‘recognizably
correct possible description.’

The rest of this lecture will be devoted to two tasks. I shall try to develop
some further rewards of the analysis so far assembled, some consequences it
throws off, and to show also how it is that the two sentences (“The baby cried.
The mommy picked it up”) can warrantably be said to be from ‘a story.’ I
start with the latter task.

It ought to be apparent that, that the children whose talk is reported in
Children Tell Stories were asked to tell a story, is not definitive of their having
done so. It is, for one, at least possible that the younger ones among them are
not capable of building stories, of building talk that is recognizable as ‘a story’
or, at least, as ‘a possible story.’

It happens to be correct, for Western literature, that if some piece of talk
is a possible description it is also, and thereby, a possible story or story-part.
It appears, therefore, that having established that the two sentences are a
possible description I have also, and thereby, established that they possibly are
(at least part of) a story. To stop now would, however, involve ignoring some
story-relevant aspect of the given sentences which are both interesting and
subjectable to analysis. So, I go on.

Certain characteristics are quite distinctive to stories. For example, there are
characteristic endings; ‘and they lived happily ever after,’ and characteristic
beginnings; ‘once upon a time.’ I shall consider whether the possible story, a
fragment of which we have been investigating, can be said (and I mean here,
as throughout, ‘warrantably said’) to close with what is recognizable as ‘an
ending,’ and to start with what is recognizable as ‘a beginning.’

In suggesting a difference between ‘starts’ and ‘beginnings,’ and between
‘closes’ and ‘endings’, I am introducing a distinction which shall, as this
course of lectures proceeds, come to have some prominence. The distinction,
which is by no means original, may be developed by considering some very
simple observations.

1 A piece of talk which regularly is used to do some activity — as “Hello”
is used to do ‘greetings’ - may not invariably be so used, but may do other
activities as well - as “Hello” is used to check out whether another with
whom one is talking on the phone is still there or has been cut off - where it
is, in part its occurrence in ‘the middle’ and not ‘the start’ of a conversation
which serves to discriminate the use being made of it.

2 Certain activities not only have regular places in some sequence
where they do get done, but may, if their means of being done is not found

262

Part III

there, be said, by Members, to not have occurred, to be absent.

For example, the absence of a greeting may be noticed as the following
data, from field observations, indicates. The scene involved two adult women,
one the mother of two children, ages six and ten. The kids enter, and the
following ensues.

Lady: Hi

Boy: Hi

Lady: Hi, Annie

Mother: Annie, didn’t you hear someone say hello to you?

Lady: Oh, that’s okay, she smiled hello.

Mother: You know you’re supposed to greet someone, don’t you?

Annie: ((hangs head)) Hello.

3 Certain activities can only be done at certain places in a sequence. For
example, a third strike can only be thrown by a pitcher after he has two strikes
on a batter.

Observations such as these lead to a distinction between a ‘slot’ or ‘position’
and the ‘items’ which fill it, and to proposing that certain activities are
accomplished by a combination of some item and some slot.

The notion of ‘position’ or ‘slot’ serves for the sociologist to mark a class
of relevance rules. Thus, if it can be said that for some assertable sequence
there is a position in which one or more activities properly occur, or occur if
they are to get done, then: The observability of either the occurrence or the
non-occurrence of those activities may be claimed by reference to having
looked to the position and determined whether what occurs in it is a way of
doing the activity.

An instance of the class of relevance rules might run: To see whether a
conversation included ‘greetings,’ look to the first utterance of either party and
see whether there occurs in it any item which passes as 'a greeting,’ items such
as (hello, hi, hi there, etc.). The fact that the list contains the term ‘etc.’ might
be deeply troublesome were it not the case that while we are unable to list all
the members of the class ‘greeting items’ we can say that the class is bounded,
and that there are some utterables which are not members of it, perhaps, for
example, “Colorless green ideas sleep furiously.” If that and only that
occurred in a first utterance we might feel assured in saying that a greeting did
not occur.

Consider just one way that this class of relevance rules are important. First,
roughly; they permit the sociologist to non-trivially assert that something was
absent. Non-trivial talk of an absence requires that some means be available
for showing both the relevance of occurrence of the activity that is proposedly
absent, and the locating of where it should be looked for to see that it did not
occur. Lacking these, an indefinite set of other activities might equally well be
asserted to be absent given some occurrence, and the assertion in question not
being discriminable from the (other) members of that indefinite set, it is
trivialized.

hectare 2 (R)

263

It does seem that for stories it is correct to say that they can have
beginnings, and we can then inspect the items that occur at their start to see
whether they can be seen to make ‘a beginning.’ Given, further, that stories
can have endings, we can inspect the items that occur at their close to see
whether they can be seen to make ‘an ending.’

While my main interest will be with the story’s start as a possible proper
beginning, let me briefly consider its close: “She went to sleep.” With this, the
speaker would seem to be not merely closing, but closing by making a proper
ending. It so seems by virtue of the fact that such a sentence reports an
occurrence, or can be heard as reporting an occurrence which is a proper
ending to something for which endings are relevant and standardized, that
very regularly used unit of orientation, the day. A day being recognized as
ending for some person when they go to sleep, so a story may be recognized
as closing with an ending if at its close there is a report of the protagonist’s
having gone to sleep. This particular sort of ending is of course not at all
particular to stories constructed by young children; it, and other endings like
it, from ‘the last sleep,’ death, unto the shutting down of the world, are
regular components of far more sophisticated ventures in Western literature.

Let me turn then to the start, to consider whether it can be said to be ‘a
beginning.’ I shall attempt to show that starting to talk to adults is, for small
children, a rather special matter. I shall do that by focussing on a most
characteristic way that small children, of around the age of the teller of the
given story, open their talk to adults, i.e., the use of such items as ‘You know
what?’ I shall offer an analysis of that mode of starting off, which will
characterize the problems such a start can be seen to operate as a methodical
solution to.

The promised analysis will warrant my assertion that starting to talk is, for
small children, a special matter. That having been established, I shall turn to
see whether the particular start we have for this story may be seen as another
type of solution to the same problem that I will have shown to be relevant.

If I can then show that another solution is employed in our problematic
utterance (the sentence “The baby cried”), I will have shown that the story
starts with something that is properly a beginning, and that therefore, both
start and close are ‘proper’ beginning and end. Such, in any event, are my
intentions.

I begin: First, roughly, and only as an assumption (though naively, the
matter is obvious) I assert that kids have restricted rights to talk. That being
the case, by assumption, I want to see whether the ways that they go about
starting to talk, with adults, can be most adequately seen to be solutions to the
problem which focusses on needing to have a good start if one is going to get
further than that. Starts which have that character can then be called
‘beginnings.’

Now, kids around the age of three go through a period when some of them
have an almost universal way of beginning any piece of talk they make to
adults. They use things like: “You know what, daddy?” or “You know
something, mommy?”

264

Part III

I will introduce a few rules of conversational sequencing. I will do that
without presenting data now, but the facts are so obvious that you can check
them out easily enough for yourself; you know the rules anyway. The
sequencing rules are for two-party conversation; and, since two-party conver-
sation is a special phenomenon, what I say is not intended as applying for
three-or-more party conversation.

One basic rule of two-party conversation concerns a pair of objects,
questions and answers. It runs: If one party asks a question, when the question
is complete, the other party properly speaks, and properly offers an answer to
the question, and says no more than that. The rule will need considerable
explication, but for now it will do as it stands.

A second rule, and it’s quite a fundamental one, because by reference to it
the infinite character of a conversation can be seen is: A person who has asked
a question can talk again; has, as we may put it, ‘a reserved right to talk
again,’ after the one to whom he has addressed the question speaks. And , in
using the reserved right, he can ask a question. I call this rule ‘the chaining
rule’ and in combination with the first it provides for the occurrence of an
indefinitely long conversation of the form Q, A, Q, A, Q, A, Q, A. . .

Now, the characteristic opener that we are considering is a question (e.g.,
“You know what?”). Having begun in that way a user who did not have
restricted rights to talk would be in a position of generating an indefinite set
of further questions as each question was replied to, or as the other otherwise
spoke on the completion of some question.

But the question we begin with is a rather curious one, in that it is one of
those fairly but not exceptionally rare questions which have as their answer,
another question, in this case the proper and recurrent answer is “What?” The
use of initial questions of this sort has a variety of consequences. First, if a
question which has another question as its proper answer is used and is
properly replied to, i.e., is replied to with the proper question, then the
chaining rule is turned around, i.e., it is the initial answerer, and not the initial
questioner who now has the reserved rights to speak again after the other
speaks. The initial questioner has by his question either not assumed that he
can use the chaining rule or has chosen not to. (Note that we are not saying
that he has not chosen to invoke the chaining rule; he has instead given the
choice of invoking it to the initial answerer. There are two different
possibilities involved.)

Second, the initial questioner does not only make his second speech by
virtue of the chaining rule but he makes it by virtue of the first sequencing
rule i.e., by reference to the fact that a person who has been asked a question,
properly speaks and properly replies to it. His second speech is then not
merely not made as a matter of either the chaining rule or his choice by some
other means of making a second speech, but it is something he makes by
obligation, given the fact that he has been asked a question, and is therefore
obliged to answer.

Third, however, the question he is obliged to answer is ‘an open one’ in the
sense that what it is that an answer would be is something that its asker does

Lecture 2 (R)

265

not know, and further, is one that its answerer by the prior sequence does.
What an answer is, then, to the second question, is whatever it is the kid takes
to be an answer, and he is thereby provided with the opportunity to say
whatever it is he wanted to say in the first place. However, now, not on his
own say so, but as a matter of obligation.

In that case, then, and the foregoing being a method whereby the
production of the question “You know what?” may be explicated, we may
take it that kids take it that they have ‘restricted rights’ which consist of a
right to begin, to make a first statement and not much more. Thereafter they
proceed only if requested to.

And if that is their situation as they see it, they surely have evolved a nice
solution to it.

With the foregoing we can say then that a focus on the way kids begin talk
is appropriate, and we can see whether the beginnings of stories - if they are
not made of the culturally standardized beginnings (such as ‘Once upon a
time’) - might be seen to be beginnings by virtue of the special situation
which kids have vis-a-vis beginning to talk.

We may arrive at the status of “The baby cried” as a proper beginning, in
particular as a start that is a beginning by virtue of being a proper opener for
one who has restricted rights to talk, by proceeding in the following way. Let
us consider another solution to the problem of starting talk under restricted
rights. I’ll begin by introducing a term, ‘ticket.’ I can show you what I mean
to point to with the term by a hypothetical example. Suppose two adults are
co-present and lack rights to talk to each other; for example, they have never
been introduced, or whatever. For any such two persons there are conditions
under which one can begin to talk to the other. And that those conditions are
the conditions used to, in fact, begin talk is something which can be shown
via a first piece of talk. Where that is done, we will say that talk is begun with
a ticket. That is, the item used to begin talk is an item which, rights not
otherwise existing, serves to warrant one having begun to talk. For example,
one turns to the other and says, “Your pants are on fire.” It is not just any
opening, but an opening which tells why it is that one has breached the correct
silence, which warrants one having spoken. Tickets, then, are items specially
usable as first items in talk by one who has restricted rights to talk to another.
And the most prototypical class of tickets are ‘announcements of trouble
relevant to the other.’

Now it is clear enough (cf. the discussion of norms earlier) that the
occurrence of a baby crying is the occurrence of a piece of trouble relevant to
some person, e.g., the mother of the baby. One who hears it gains a right to
talk, i.e., to announce the fact that it has occurred, and can most efficiently
speak via a ticket, i.e., “The baby cried.” That being so, we can see then that
the opener “The baby cried” is a proper beginning, i.e., it is something which
can serve as a beginning for someone whose rights to talk are in the first
instance restricted.

With the foregoing we have established that the story we have been
examining has both a proper beginning and a proper end, and is thus not only

266

Part III

a story by virtue of being a possible description, but also by virtue of its
employing as parts, items which occur in positions that permit one to see that
the user knows that stories have such positions, and that there are certain
items which, when used in them, are satisfactory incumbents.

Lecture 3

‘Everyone has to lie ’

There is no transcript of this lecture. Its materials became the paper ‘Everyone has
to lie,’ now in M. Sanches and B. G. Blount ( eds ), Sociocultural Dimensions of
Language Use (New York: Academic Press, 1975), 57—80. For the materials
in lecture form, see Spring 1 96 7 , lectures 8 and 9.

267

A note on the Group Therapy
Session materials

The bulk of the lectures of Spring 1966 deal with a segment of talk in a group
therapy session for teenagers. The segment had been introduced in June 1 965 as
a final exam, and a series of lectures dealing with the materials was given in
Tall 1965- The next presentation, given in Fall 1966, is not included in this
volume. In the main, the Fall 1966 lectures were a reorganized repetition of the
Spring 1966 lectures. There were some new considerations , expansions, etc., and
many of those have been incorporated into the Spring 1966 series.

Whatever introductory discussion Sacks offered was not transcribed in the
Spring or Fall 1966 lectures. In its place is a brief introductory note, made up
of Sacks’ remarks about the June 1965 final exam, and an introduction to the
materials when he first presented them in Fall 1965.

Introduction

It seems to me that the core of this 1 1 -minute segment from a group therapy
session for teenage kids constitutes an ‘initiation ceremony.’ As such, among
the things in it that are of interest, is the fact that it isn’t done via a script. The
participants are a therapist, male, about 35, three current patients, male,
between 16-18 years, and the new entrant, another teenaged male. One of
the kids in the group knew beforehand that somebody new was going to come
that day, but that’s about it. We are not dealing with a ‘scripted’ ceremony,
nor with one for which the various parts have, even roughly, been previously
assigned or previously worked out. And yet there are certainly features in it
which the literature on initiation ceremonies proposes to be present in such
events.

But our task is not going to be so much to see the extent to which that is
so, but to try to come to terms with the conversation itself, as best we can.
Now, the quickest way to find ourselves feeling hopeless at the task is to
attempt to deal with the whole thing at once. We’re going to be picking at
it piece by piece. And there are a lot of events in it that are, to the altogether
naive eye, quite remarkable. That is, without any analysis and simply by
inspection you can find some things that you might take to be worth thinking
about. Probably the most striking thing right off, would be that section on
page 3 of the handout, where there’s a series of questions run off in about 45
seconds, by the given members of the group to the new entrant, in which they
ask him things about himself, giving in each of the questions what they take
to be the answer. And the questions have some rather extraordinary detail in

268

Note

269

them; for example, question ten, “you’ve been kicked out of school once.” It
is quite obvious that how these guys have that information, and what they do
with it, are issues. And there are, aside from that, a variety of other things
involved in this segment that we’re going to have to consider in one way or
another.

Handout: Group Therapy Session

Segment

0:00 Therapist'.
Roger.

Therapist'.

Roger.

Al :

Roger:

Therapist :
Al:

Roger:

Therapist:

Al:

Roger:

Al:

0:30

Roger:

Al:

Roger:

Therapist:

Roger:

Al:

1:00 Roger:

heh heh

Kids don’t drive long. They start off when they’re si-by
the time- when they’re sixteen, by the time they’re
eighteen they’re back walk(hh)ing hehh through
circumsta(hh)nces hehhehh
beyond their control.

Ehhehh yes hehhehh hotrodders // never-
It’s all the cops’ fault.

Now the Soshes keep driving. Daddy’s car and
everything you know,

It’s the cops’ fault.

Yeah it’s not ours we’re per- we’re / / merely
It’s the insurance companies and the cops.

Yeah,

Damn right.

Kids are completely innocent heh

We do race but the times we do get tickets are the only
times when we maybe goin’ a long distance, uhh I dunno,
1- depends on how many accidents you have, and how
many of these tickets you get.

I have got a speeding ticket on the average of one every
three and a half months. Since I’ve been driving.

I haven’t- I have never gotten a // ticket.

And I deserve it.
ha ha ha

I deserve it. Because I always drive about seven miles an
hour faster than everybody else. You know? Like
everybody’s driving thirty five I’m driving about forty
two somethin’ like that. So this is the conseque(hh)ence I
have to pay. ‘Cause I can’t stand driving that slow.

I have— I’ve- I’ve— all the time / / I’ve been driving I
haven’t gotten a ticket but I’ve always- but I’ve always-
Specially when I think I can handle a car at that speed. In
traffic.

This is the text of the original Spring 1 966 transcript, with an inclusion from
retranscribing done in Fall 1966. With this new transcript came a new set of names
for the participants: Mel, Joe, and Henry are now Al, Ken, and Roger, respectively.

270

271

Handout: Group Therapy Session Segment

Ah

Therapist:
Ken :

( ):

Al:

Kerr.

Roger.
rKen :

1 Roger:

2:30

Ah.

Roger:

Ken:

Roger:

3:00 Jim:

Therapist:

Jim:

Therapist:

Jim:

Therapist:

Jim:

Ken:

Therapist:

Roger:

Jim:

Therapist:

( ):

Ken:
Roger:

( ):

I shoulda gotten a ticket // but uh
Uh huh

Oh I can think of a million times when I should’ve gotten
a ticket, going on the freeway. I’m always goin’ seventy
five. I mean I’m just I’m not racing, I’m right along
with- or I’m right behind the guy ahead of me and he’s
just he’s goin’ as fast as I am and lotta people pass us up
like we’re sittin’ still.

Oh yeh.

Like Roger.

And the cops- the cops don’t look at you when you
just-

But they’re Stockers.

If you’re in a jeep-

But you drive a Stocker. When a cop sees a hopped up
car, he don’t care if you’re going forty five you must be
doin’ somethin’ wrong, and if he wants to be mean, he
can bust you on a thousand things.

He doesn’t have to have a reason, you know you’re not s-
cops aren’t supposed to talk to you unle- s-stop you
unless you’re doing something suspicious. If you have a
hotrod that’s enough.

He can bust ya on a thousand things non-moving
violations if you give him any lip, even if you didn’t do
anything wrong. They can just go over your car and write
you up ticket this long.

Sure. Well let him. You know-

On Hollywood Boulevard the other night they were
giving tickets for dirty windshields.

((door opens))
hh

Hi, Jim / / c’mon in.

H’ wary a

Jim, this is uh Al,

Hi

Ken,

Hi

Hi

Roger.

Hi

Hi

Jim Reed.

((cough))

We were in an automobile discussion
discussing the psychological motives for
hhhhhhh ((not laughter, heavy breathing))

272

3:30

4:00

4:30

Al:

Ken:

( ):
Ken:

Al:

Roger:

( ):
Ken:

Al:

Ken:

Roger:

( ):

Ken:
Roger:
Al:

Roger:
Al:
r Ken:
l Al:

( ):

Al:
Roger:
Al:

Ken:

Al:

Roger:

Al:

Ken:

Al:

Part 111

drag racing on the street.

I still say though that if you take if you take uh a big
fancy car out on the road and you’re hotroddin’ around,
you’re- you’re bound to get you’re bound to get caught,
and you’re bound to get shafted.

( , //, )

W 1 look- I m gonna-

Not unless ya do it right. / / That’s the challenge- that’s
the challenge you wanna try and do it right so you don’t
get caught!

That’s the problem with society. Hehheh
(Ya gotta-)

In that Bonneville of mine. I could take that thing out
and if I’ve got a tie and a- a sweater on and I look clean?
ninety nine percent of the time a guy could pull up to me
in the same car, same color, same year the whole bit, roar
up his pipes and he’s in a dirty grubby tee shirt, and the
guy’ll pick the guy up in the dirty grubby tee shirt before
uh he’ll pick me up.
hah

Ju- just for-

But not many people get picked up in a Pontiac station
wagon.

hhhh Now (well) I agree it looks like a Daddy- it looks
like a damn // Mommy’s car.

Ken, face it. You’re a poor little rich kid.

((cough))

Yes Mommy/ / thank you.

Face the music.

Ok. Now you’ve got that er outta yer system. Now
you’re a poor little rich kid we’ve told you that.

And we also decided you’re a chicken shit,
heh hehhhh

I decided that years ago the hell with ya.

( )

hhh

Now let’s see what else can we decide about // you.
Hey don’t tear him down.

We’re- Ok.

I’ve been / / torn down for ( )

< // >

We got company heh

(hh) Oh Ok.

( ) hhhhh

Tell us all about yourself so we can find something bad
about you.

Ken :

Handout: Group Therapy Session Segment 273

Yes, hurry up.

Roger:

Well first of all you must be crazy or you wouldn’t be

Ken:

here,
heh heh

Jim:

Yeah that’s how it goes.

Al:

Secondly, you must be an under-achiever.

Jim:

Yeah,

Roger:

You hate your mom and dad huh

Al:

—third of all. ((tied back))

Jim:

Oh, sometimes.

Al:

Fourth, you like to drive cars fast.

Jim:

Yeah,

Al:

Fifth you like to ta- you like uh wild times,

Jim:

Mm hm,

Ken:

He smokes like / / me, see?

Al:

Sixth you like booze,

Jim:

Huh? // Yeah

Al:

Seven you like to smoke

Jim:

Mm hm

5:00

Roger:

And seven you’ve been arrested for rape and other things.

Al:

Eight you heh

Jim:

No, not that.

- Ken:

hehheh

-Jim:

hehh you’re gonna ta(hh)lk

L Al:

Eight you- Eight you uhm give lip back to everybody,

Jim:

Yeah,

Al:

Nine, you uhh cut classes,

Jim:

Yeah,

Al:

Ten, you’ve been kicked outta school once,

Jim:

Yeah,

Al:

Eleven, uhhh

( ):

wshhhh

Al:

hehh hhh hehh

Roger:

We’re doin better than he is hahhehhehh Proceed Em

r Al:
i Ken:

(w’ya) I’m with ya.
heh heh Your turn.

(Oh-)

Roger:

Sure. Are you just agreein’ because you feel you want to

Jim:

uh . . .
Hm?

Roger:

You just agreeing?

Jim:

What the hell’s that

5:30

Al:

It’s- / / agreeing?

Roger:

Agreeing.

Jim:

Agreen?

Roger:

Yeah.

274

Part III

6:00

r Al:

[ Ken:

Jim\

( ):

Roger:

( ):

Several:
Jim:
rKen:

l Al:

Jim:

Ken:

Jim:

Roger:

( ):

Jim:

Roger:

Jim:

Roger:

Jim:

Roger:

Al:

Roger:

Ken:

Roger:

Ken:

Roger:

Ken:

Roger:

Ken:

Roger:

Ken:

Roger:

Ken:

6.30

[

(

Al:

(

Jim:

Ken:

)■

[

Al:

Ken:

With us. Just going along with us.

Agreein’.

No.

(Yeh)

Saying / / yes, yes, hehh hehh hehh
((cough))

((laughter))

Well it’s true. Everything he said is true, so . . .

Have you been kicked out of school?

You wouldn’t be in the group if you didn’t.

Yeh

What school did you go to?

I went to Palisades.

Went?

Oh oh

Yeah. And then to University from there.

Oh I went to Uni, I graduated there.

You did?

Bitchin’ school isn’t it.

Kinda groovy

( )

What’d you do with the match // ( )

You don’t like it do you?

I got it.

Here you go chum.

EVERYBODY’S FASTER THAN ME TODAY!

You got reflexes like a slow turtle.

God damn you hehhahh

Remind me later I’ll show you a little game about your
reflexes.

Hey you got another one of those Santa Fe’s?

Nah it’s the last one.

Last one?

Pity hehheh
heh heh heh

Usually there’s a broad in here. / / Pretty (tough)

Her name is Louise. She’s pretty good, she’s real nice,
((cough))

I’d like to ( )

There’s a lotta her to love, // Hm?

There ’s—

Sound like an old man.

Tchhhh hehh This is the old- this is the- the couch
society of- what is it?

Wanna join our army?

( )

275

Handout: Group Therapy Session Segment

Jim :
Al:

7:00

Ken:

Al:

Roger:

Ken:

Roger:

Al:

Roger:

Al:

Ken:

Al:

Ken:

Ken:

Roger:

Al:
r Jim:

1 Roger:
Therapist:
Roger:

Jim:

Roger:

Ken:

( )
r Ken:
[ Al:

Roger:

Ken:

Roger:

Jim:

Roger.

Ken:

Roger:

Jim:

Al:

Ken:

Al:

Roger:

Jim:

Huh?

Wanna join our army? We’re gonna go defeat the king of
Greece. 1
We’re gonna-

The Academic Counseling Service army’s gonna go-
Alright Al, / / he’s been in here ten minutes.

( ) Hey your match is still goin’, Al.

Al, he’s been in here ten minutes let’s judge his character.
Ok. Ready Roger?

Ok.

You’re first.

You’re gonna get killed,
hhehh

Keep your mouth shut and don’t say a damn wo(hh)rd
heh

He’s not at all like Ken, (first)

No

Who’s Ken?

He’s more like uh Al and I.

That’s Ken. (And that’s Al.)

’s what you refer to as a hippie hheh He’s been in it up
to his neck,

Yeh

coupla times,
heh heh
ough

Taste good?

I- no I bit off the end of it I was chewing on the end of
it.

uhhm

Bitchin!

Havin, a big hassle with your folks.

Yeah,

Right. Daddy wants to keep you down.

((cough))

Mommy?

Hm?

Good. Mommy you can tell what to do.

((cough)) Goddam cancer.

See I told you. We know everything about this guy
a(hh)lready.

Ya have any brothers or sisters?

No.

1 Some 30 minutes earlier, Al has said: “We oughta go to someplace like Greece
where they have a king y’know, an’ challenge ’em with- like the Academic
Counseling Center’s army y’know, versus their army y’know,”

276

Part III

Roger:

Jim:

Roger:
7:30 Jim:

Al:

Roger:

Jim:

Ken:

Al:

Roger:

Jim:

Roger:

Jim:

Al:

Roger:

Jim:

Roger:

( ):
Jim:

Al:

Ken:

Roger:

Jim:

Ken:

Jim:

Al:

Roger:

Al:

Jim:

8:00

Al:

( ); .
Therapist:
'Roger:

-Therapist:

.Ken:

Al:

Roger:
8:30 Al:

Jim:

Ken:

Roger:

Jim:

Daddy sorta sheltered you.

No, not really.

No?

Not really.

Y’ sorta sheltered them.

Ya have a car?

No.

((gasp))

Uh oh. One point against / / you.

You hadda car?

Nope.

You want a car?

Oh hell yes hehh

You’ve stolen / / a car

Daddy doesn’t want you to have one.

No, he wants me to have one, my old lady doesn’t heh

Oh! Main conflict with mother.

whhhhhh

Yeh

heheh

hehheh his

Mother is too dominating.

Yeah.

But you been had already fella hehh

Mm hm

Ok.

Bring your mother in we’ll work on her hahh

hehhehhhh

Ok heh

((pause))

What are you thinking, old man?
heh heh

hahahh I’m thinking Jim has certainly had it. Heh
Whip out your pad— Now, now, now- whip out your
pad and write down what we say. Hehheh Take dictation.
(He hasn’t been here) five minutes and you guys-
You gotta pardon these guys they’re all they’re all the uh
/ / social- social outcasts of the world.

I’m vicious- I’m vicious, he’s vicious-

This one’s chicken shit, he’s a bastard, I’m a mau mau.

Ok good. Now we’re all settled, what are you.

(Well I-)

You a hood?

This is an abnormal session, / / see we’re not together
without the broad.

Yeah.

277

Jim :

Al:

Ken :

Roger:

Ken :

Jim:

Roger:

Ken:

Al:

Ken:

Al:

Ken:

Roger:

Ken:

Al:

Roger:

Jim:

Al:

9:00 Roger:
Ken:
Roger:
Jim:
Jim:
Al:
Roger:
Al:

Roger:

Al:

r( ):

-( ):

Roger:

Al:
Roger:
Ken:
Roger:
Jim:
9.30 Ken:

Handout: Group Therapy Session Segment
Yeah.

See we gotta have the broad here cause she- she unites us.

heh heh She- she / / unites us yeah

She sorta keeps us on the road to some // uh meaning.

She keeps us on the road heh table’s always ready and heh

heh heh

Oh well

heh heh

Oh well you know that’s the way things go,
heheh

That’s the way it 1- that’s the way things lie down,

y’know

Hey-

He sorta keeps his mouth shut and writes down things
whenever you say something important.

Yeah.

And- and then if you ask him to do something // you
have to pay him though.

He’s a good guy, though.

Ok.

Y’know

Eventually you’ll become sane,
heh Yeah which is-

Or your mother will, which is // not what you think.
Yeah.

I hope so.

But uh

You’re not / / crazy, it’s your mother huh

It takes about- I’ve been here for four months you know

and it’s- it’s getting better now I used to think that I was

a bird’n used to fly you know but

Yeah he used to walk out on us y’know he thought he

was above us.

Yeah but now I’m— now I’m below you.

hehh

heh heh

See I corre- I corre- I corrected that fault. Hehhehh I
gave him an inferiority complex.

And I got him to shave.

Yeah

ehhehh

I’m not grubby or nothin’

N(hh)o

heh Hey this is- this is the Academic Counseling Service
it’s called the family- family circle.

278

Part III

Roger.

Ah.

Ah

Roger :

Kerr.

Roger:

- Several :

Jim :

Several :

Roger:

Jim:

Jim:

Ken:

Ah.

Roger:

( ):

10:00 Ken:

Ah

Jim:

Ken:

Roger:
r Ken:
l Ah
Roger:
r Ken:
'-Ah.

Ken:

Jim:

Ah

Ken:

Roger:

( ):

( ):

Roger:

( ):

Ah.

10:30 Roger:
Ken:

It’s not really a- an Academic Counseling Service / / it’s
sort of a drive-in nuthouse, hehh
Then your father comes in here-

hehhhhehhahaha Then you rent you- then you rent a
counselor. Five dollars an hour and fifty cents a question.
Don’t consider this one but on the whole you’ll enjoy
’em. Fact it’s a good way to spend Saturday morning,
heh hehh
If you’re stupid.

After Friday night.

The morning / / after the night before.

((laughter))

Yeh

((laughter))

After a coupla / / cups of coffee you’re alright.

( )

Yeah.

Couple dozen cups of coffee.

Chew a cigar now and then, and . . .

I don’t smoke,
whhhhh

( )

(He just / / started last night.)

-just eat it.

You gotta believe- you gotta believe it we’re the only two
in here that smoke other than him.

Doesn’t Lu smoke?

No,

No she doesn’t smoke. She said she used to.

She’s a good gi(hh)rl ahhhaha

She- even-

(She like she eats it)

She smokes cigars that’s about all,

I see.

They smoke, I chew ’em.

He eats it.

’s well chewing ’em’s half the fun (I belie-)

hehh

heh

Uh oral gratification or something heh
hehh

Let’s see what else can we tell / / him ( )

They say that’s why bubble gum and that shit’s good- uh
so popular y’know
Why?

Roger :

( ):
Roger:
Several:
Roger:

Jim:

Al:

Roger:

Several:

Roger:

Several:

Al:

Roger:

Al:

Roger:

11:00 Al:

Roger:

Ken:

Roger:

Al:

Roger:

Al:

Roger:

Roger:

Al:

Roger:

Ken:

Jim:

( ):
1 1:30 j-A/:

Roger:

Jim:

Handout: Group Therapy Session Segment 279

There’s a trend to uh put things in your mouth like

thumbs, y’know

heh

No really.

((laughter))

No! Well why do you think babies // suck their thumb
and adults chew gum and smoke everybody wants to
keep something in their mouth.

(Mmhh)

Well you know // ( )

No, I’m not talking about eating or nothing, and I’m not
talking about what you’re thinking of either
((laughter))

No really I read it in some book y’know
((laughter))

((cough)) Oh c’mon // aheh hehh
It’s really the truth!

One more // thing

Look at im. Look at im suckin on that pipe all goddam
morning hehh

He sucks on his dentures too! hehh Now listen one- one
more thing. There’s people watching us now. They got a
tape recorder and they record // (everything)

(That is the) microphone. Occasionally there’s somebody
in there judging us. They’re baby headshrinkers. They’re
learning how to be.

Don’t worry / / about ’em they destroy the tapes / /
afterwards.

-what he is.

He’s not quite one. He will be one in two months,
’sright.

He’s writing a // (whatchacallit) article.

(a paper)

(But today-) Yeah. But today we’re not being uh viewed.
We’re just being taped.

No we’re being viewed there’s somebody in there cause
they lit a cigarette.

But he’s blind.

Hey don’t worry about it,

How c- how c’d you see him?

(Well )

It’s- can see us, but you can’t see-

That’s not a mirror. // When you turn the lights on

y’know

The see-through kind? / / Yeah.

280

Part III

Roger: Mh. See it’s darker in there. //I think.

Jim: Uh huh

Ken: Wave at him he’s a nice guy.

r Al: Anything else, Roger?

'-Ken: They give parties.

Roger: And if it gets private we can always shut the curtain.

( ): heh heh

Jim: heh great.

Al: Anything else Roger that you think we oughta let him

on- in on, or tell him?

Roger: Nah

Ken: We’re all nuts?

Roger: Let him slide for a few weeks.

Al: (hh)Ok.

Roger: Get the hang of things hehhehh

Al: Then we’ll really go after ya. Then when you know when

we get-

Roger: We’ll let you rest with your problems now.

12:00 Al: We’ve told ya all your problems now you gotta rest with

’em then / / we’ll then we’ll then we’ll tell you how to
solve them / / and you’ll pay us off.

Jim: Yeah.

Roger: You gotta get warmed up to it

( ): ((cough))

( ): Yeah.

Roger: and vice versa.

Lecture 04. a
An introduction sequence

The data that we’ll be by and large considering in this lecture can be found
in the middle of the second page of the handout. 1

Roger

: On Hollywood Boulevard the other night they were giving tickets

for dirty windshields, ((door opens))

Jim-.

hh

They.

Hi, Jim / / c’mon in.

Jim:

H’warya

They.

Jim, this is uh Al,

Jim\

Hi

They.

Ken,

Jim:

Hi

Ken :

Hi

They.

Roger.

Roger

: Hi

Jim :

Hi

They.

Jim Reed.

Now we’ll start to pick away at this a little bit. We can start out by noticing
that the sequence in which the introductions are done is not, itself, a disorderly

fact, i.e.,

it is not a random matter. There is probably only one crucial place

in the determination of the sequence in which the introductions will be done,
and that crucial place may well be noticed via the “uh” in “Jim, this is uh

Al.”

You might consider, if introductions are going to take place, and the
therapist is going to do them, 2 what the import would be of various ways that

Lectures 04.a and 04. b consist of two lectures given in Fall 1966 (lectures 1 and
2) and again in Spring 1967 (lectures 6 and 7). They deal with matters that are not
discussed in the Spring 1966 lectures, but they have been placed where they would
belong in the sort of progression through the data that Sacks used for the Spring
1966 lectures. While the discussion contained in lectures 04. a and 04. b is misplaced,
it is not altogether anachronistic, since the later Fall 1966 discussion is not the first
mention of these matters. There is a brief consideration in Fall 1965, lecture 3, at
p. 148

1 In Spring 1966 a new transcript was produced, and the names Mel, Joe, and
Henry were changed to Al, Ken, and Roger, respectively.

2 The unedited segment runs off differently: “You might consider, if introductions
are going to take place, and this fellow Dan is going to do them. . [emphasis

281

282

Part III

he would go through the three persons he’s going to introduce Jim to.
Obviously there’s more than one way he could proceed through those three
persons.

You could ask, for example, if we knew the first that he had done, could
we then predict the second and third? And in such a way, now, not that we
would be right, but where, if we weren’t right then something of interest
would be present. That is, either we’d be right, or if we weren’t, then we
might well be on our way to explaining, not that we were wrong, but that
something strange took place. Such a question is possibly relevant, insofar as
for the participants to the ‘introduction sequence,’ while perhaps they
couldn’t predict who would be introduced first of the three, they could feel
assured, given the first, if they hadn’t been picked then they would be second
or third; for example, be ‘prepared’ to be second. We’re talking, then, of
some way of proceeding through what would be seen by them to be a
sequence which, while it might have alternative ways of beginning, given a
beginning it would then have correct ways of proceeding. And, perhaps not
any beginning would be correct.

It’s also the case that not only should that sequence go in that order, but
it should go to completion. Introductions go in rounds. That is, not only
might there be the possibility to predict whether one will be second or third,
but one might predict that there would be a second and third, given a first.
The rule that said ‘do an introduction here’ would also say ‘do a round of
introductions here.’

The idea is to see whether there is some oriented-to orderliness in that
actual sequence. And in this case it would seem that the ecological distribution
in the room presents proper alternative sequences. That is, the seating is
arranged in a rough semicircle with the therapist closest to the door, then Al,
then Ken, then Roger. When Jim is introduced, he is standing near the door.
If the therapist begins with Al, he would expectably proceed to Ken, and then
Roger. But not, e.g., Al, Roger, Ken. And if he begins with Roger, he’ll go
Roger, Ken, Al, but not Roger, Al, Ken. Further, he wouldn’t begin with
Ken. In that regard, then, it might be that ‘Al, Ken, Roger’ or ‘Roger, Ken,
Al’ would be equivalent, but all of the other possibilities would be different.

added]. In contrast to the pronominal changes discussed earlier (Fall 1964-Spring
1965, lecture 2, p. 18, n. 1 and Appendix A) as matters of policy, the change from
“this fellow Dan’’ to “the therapist” runs acutely counter to editorial policy. The
policy was to preserve just such references, as is, in line with Sacks’ deep concern with
the issue of category-selection. This lapse from editorial policy occurred in response
to an anachronism which cropped up in the editing process. Most roughly, text from
a later lecture was introduced into a set of earlier lectures, the earlier lectures having
been produced by reference to a transcript designating this speaker ‘Therapist,’ that
designation having been introduced by Sacks in his 1965 transcript. The text in
question, however, had been produced by reference to a new transcript designating
this speaker ‘Dan.’ The incongruity was resolved by using the term appropriate to the
earlier lecture series and its transcript, i.e., ‘the therapist.’ For a more elaborate
discussion, see Appendix B.

Lecture 04.a

283

Any one of these alternative possibilities might be informative about the set
of persons; e.g., if one wanted to point up the fact that, say, Ken was a
distinguished person in one or another way, then one might well begin with
him. That is, the sequence is usable to present various kinds of information
to the recipient, and to the other parties. But the activity can also be done in
such a way that no such information is present, or considered to be present or
to be used.

If the series goes ‘Al, Ken, Roger’ then we might suppose if they were
seated in other ways than that, e.g., Ken, Al, Roger, then the introduction
would go ‘Ken, Al, Roger.’ That is, the actual order seems to be partially
independent of who happens to be seated in each position. Rather, they are
seated in positions relative to each other which can be used to produce an
observably orderly sequence. If the order ‘Al, Ken, Roger’ is used, then there
is no suggestion, apparently, that Al is the most significant person present, but
parties can see that Al’s being picked first is a matter due to his having been
seated there; they can see that if Al were seated elsewhere, he wouldn’t have
been first.

Now that’s a matter that one wants to be somewhat attuned to. After all,
Al was picked first, Ken second, Roger third. That they can see that it’s not
‘Al’ who was picked first, but ‘the person who is sitting where Al happens to
be sitting,’ would be of some importance if it could be an issue that some
sequence would have another interpretation, e.g., that Al was picked first
because he was the most important.

The data, the actual talk, can bear alternative ways of being seen. It is
perfectly compatible with a view that provision for who was picked first,
second, third, was made by, e.g., the social status of the parties and not by
virtue of the ecology of the room, especially if the data were to be used by the
parties to it, to find out what, really, their positions are in this group. I take
it, however, that with the whole sequence available, one or another of these
alternatives is decidable on. That is, for the seating order ‘Al, Ken, Roger,’ if
the introduction went off, ‘Al, Roger, Ken,’ then Al might well be able to
figure that he was picked first not by virtue of his position in the room, but
by virtue of something else. That is then to say, that to determine what it is
that was operative, requires that several occurrences be seen to be ‘a sequence’
in the first instance.

One wants also to see that it may well be that adequate decisions could be
made as to what the therapist is doing with that sequence, with only that
sequence available. That is, on a single occasion and without discussion, it can
be available to participants that, e.g., ecological positioning is being used to
order the sequence, or that someone is being selected via social status, etc.

I go through this for what could be fairly obvious reasons. One of them is,
one might have the idea that the way in which some rule is to be detected is
by, say, collecting some 10, 100, 1,000 sequences of introductions and
plotting them in various ways, and if you seem to get a very good fit then you
figure there is a rule, and if there is not, then there isn’t a rule. What we have
to find is, if someone can be adequately insulted or adequately commended by

284

Part III

a single introduction, then how that could be done becomes a problem that
the data has to be handled to provide for.

Etiquette books give various rules for doing introductions and sequences of
introductions, and they seem to suggest that one can bring off ‘picking
somebody out’ or ‘not picking somebody out’ on a single occasion of doing
introductions. Persons can see that they’ve been properly selected to be
‘introduced first,’ or, for that matter, that they’ve been neglected, without
having to, e.g., collect several introduction series by the same introducer to
see, e.g., how often they came first, how often seventh, etc.

Now, there are also kinds of consistency constraints on what it is that goes
into a given sequence. For example, when persons are introduced to each
other, and say more than ‘hi,’ it’s not just that you have to make some
identification because the last person on the list made some identification, but
that the kind of identification they made is criterial for the kind that you
make. That is, roughly, you find persons sticking within a single device when
some sequence of identifications are being done. If the last person has said
where they come from, then you will say where you come from, etc. And that
suggests something that we want to watch for - the possible relevance of who
goes first on a round, in setting up what it is that will be used. That can matter
a great deal, obviously, since some people are ‘better’ at some things than
other things - they have a longer lineage, or they have a better job, or they
live in a nicer neighborhood than most other people. And that may matter as
to which they want to put into an introduction. Given the way a sequence
runs off, whether they get a chance to use their item is another issue. And that
consistency constraint holds for names, as well, to some extent. So that the fact
of a first name, nickname, etc., being used throughout the sequence can be
quite informative, and it can be even more informative if differences are
involved, i.e., if the therapist had said “Jim, this is Al, that’s Mr Kenneth
Goss, and this is Roger.”

Given the fact that the kind of name used is a rather important matter and
has a series of significances, we can stop to consider briefly a matter that I and
one colleague of mine have an enormous fascination with. And that is the sort
of thing that could be seen to be going on in the interchange:

Therapist : Hi, Jim / / c’mon in.

Jim : H’ wary a

which, from the point Jim starts up, gives us “Hi Jim,” followed by
“Howarya.” Is there anything interesting in “Howarya”?

First of all, “Hi, Jim” is a ‘greeting’ consisting of a version of ‘hello’ plus
a name. Now “Howarya” is not directly a ‘greeting,’ but it’s a ‘greeting
substitute.’ Greeting substitutes can go where greetings go. 3 And “Howarya”
can then go where greetings belong. It can go in first position or second; that
is to say, to begin off a conversation which properly begins with a greeting,

3 See Spring 1967, lecture 8, pp. 554-5.

Lecture 04.a

285

instead of saying “Hi,” you can say “Howarya.” In this case, that it is second
is of some significance. And that is because routinely a greeting accompanied
by a name is followed by a greeting accompanied by a name. “Hi Jim,” “Hi
Mac.”

Now, in certain circumstances, giving or not giving names can be a fairly
significant matter. And in this case it could well be a significant matter, in the
following way. This guy is a new member of the group. He probably knows
the therapist as Dr so-and-so. That is to say, in the prior company in which
he’s spoken to the therapist, say, with his parents or whatever else,
presumably that is the way they’ve talked. And in any event it may well be
an issue as to how it is that he’s going to address the therapist: Dr, Mr, last
name only, first name only? It’s not only an issue of their private relationship,
but it’s now right at the beginning of his presence among a bunch of other
guys. And his choice is something that they see. They presumably have some
way of dealing with that name problem. He doesn’t know what it is. If he
comes on with “Hi, Dr so-and-so” and they call him something more rough
and ready, then he may well already have been seen to be a kind of pansy of
sorts. That is to say, for somebody who is in a situation where they’re
watching somebody address another, the kind of address that’s used right off
is informative for them, about what relationship holds, about what sort this
one is, that one is, etc.

Now, for one, with “How are you,” the fact that there’s no name doesn’t
make a problem of who it’s talking to. It is directed, by virtue of coming in
response to the greeting, back to the person that did the first greeting, the
‘you’ being that person. Furthermore, as a ‘standardized item,’ ‘portable,’ an
‘idiom,’ whatever you like, ‘How are you’ is complete on its own. It does not
lack a name, which would be lacking in that sense if he just said ‘Hi’ in return.
By using “How are you,” then, instead of directly a greeting, the issue of a
name is avoided.

And it’s not merely that he doesn’t make a decision on what name to use,
but the fact that he isn’t making a decision is not apparent in any obvious
way. He hasn’t simply addressed the fellow “Hi [no name},” i.e., in that sense
not making a decision, but he’s built something which doesn’t need a name.
So, that he is avoiding the decision on a name is something that need not be
seen, and perhaps can’t warrantably be seen. This phenomenon of no-naming
people is a big issue, and a range of techniques for doing it are available.

Let’s turn to the way that this talk comes off. I’ve talked about a ‘sequence
of introductions.’ And the question is, is that just the way I’m talking about
the fact of three introductions? That is, there were three, I’ll call them a
sequence. Or is there anything about it that provides that ‘sequence’ is a more
useful way to go about noticing what’s going on than saying there were three
introductions.

Let’s look again at the language here. “Jim, this is Al, Ken, and Roger”
says the therapist. In its fashion it’s a terribly interesting piece of talk. The first
thing you want to notice is, it’s one sentence in which room is apparently left
for insertions of greetings by the various parties introduced. We don’t have

286

Part III

something like, “Jim, this is Al.” “Hi.’’ “Hi.” “Jim, this is Ken.” “Hi.”
“Hi.” “Jim,” this is Roger.” Instead, the second and third introductions are
tied to the first in quite clear ways, leaving aside the mere fact that they
happen to be in a sentence. What’s involved here is that they’re not only
involved in a sentence, but that for the second or third to be seen as
introductions, attention by the parties who are being addressed in those is
necessarily directed to what happened before. So, for the word ‘Ken’ to be
seen as an introduction of Ken to Jim, involves the fact that both Jim and Ken
will have been attending “Jim, this is Al,” to see ‘Ken’ as an addition to that.

One of the ways in which that is obviously of use to us, and something that
it took me rather a long while to get the importance of - and I can’t give it
here in full, I only want to suggest it - is that we have a multi-party
conversation going on, and a question we could ask is, what’s the status in a
multi-party conversation of persons who are not addressed by some utterance?
While persons who are not addressed directly are not ‘overhearers’ - in the
sense that in a two-party conversation a third person, not addressed, if it turns
out that he was attending, then he would be doing something wrong - what
are they? Do they have, say, rights to listen in because it’s in their presence and
they’re properly present? Or do they have a business to listen in because what’s
being done to the one being addressed is also in some sense being done to
them? Where their participation in the things that are addressed to them will
have required that they have attended what has been said to somebody else
in between utterances addressed to them.

In that regard, we can imagine in the first place that there is the question
of what a ‘group therapy session’ consists of. Is it, for example, a set of private
discussions between two persons, where, when you’re not being talked to or
you’re not talking, you shouldn’t really listen because it’s somebody else’s
troubles that are being talked about? One thing that’s going on here, then, is
an invocation of the fact that in a conversation like this, if you’re not talking
or being talked to directly, you’re not to behave as one who would be
conceived as ‘overhearing.’ But in these conversations you listen to talk which
is not addressed to you. And aparently you can be held responsible for that
talk, in the sense that it may be assumed that you know what’s gone on,
which hasn’t been between you and somebody else.

So, in this case, for Ken to see what has been done with “Ken,” involves
him in showing that he saw the first introduction, and saw that the form ‘this
is. . .’ remains available for his name to be attached. And not, for example,
that “Ken,” is something addressed to him independently of what has been
going on that was unaddressed to him, where, then, “Ken,” might be a signal
for Ken to say “What?” or things like that. The same, of course, goes for
Roger. Now Al doesn’t say “Hi.” I’m not sure whether he did some greeting,
e.g., by nodding, waving, etc., or just didn’t do a greeting and was treated as
having his time passed by the therapist who just went on, rather than, say,
waiting Al out. From the rest of the material that we have, I can say that not
giving a greeting isn’t something that’s beyond Al, but I just don’t know, and
nothing is going to be claimed on that score. In any event, the second and

Lecture 04.a

287

third names are not, except by virtue of their position as second and third in
a sequence which starts with “Jim, this is Al,” signals for an exchange of
greetings. They’re quite equivocal as that.

And in that regard there’s another thing that is, in its fashion, quite minor,
though in some ways it can turn out to be of larger significance, and that is,
the acknowledgements of the introduction are done with the term ‘Hi.’
What’s the character of ‘Hi’ in that use? That its use does acknowledge the
introduction, turns only in part on the fact that ‘Hi’ can do that; it turns also
on the fact that it’s ‘an introduction’ that’s taking place. For in the first
instance ‘Hi’ isn’t a term which would be said to be, among its predominant
uses, that item by which you acknowledge an introduction. It’s a ‘greeting’ by
and large. Though again, it’s not always only a greeting, in the sense that it’s
not merely doing a greeting in this case.

What it is that a given term is doing, is something that may well require
some analysis of its locus in conversation, to find and to argue. (In the case of
‘greetings’ it may be very simple, but I’m not sure it’s all that simple.) And
in that sense, then, the specification of the uses of sets of terms require a
formulation of the structure of conversational possibilities. And if that’s so,
then we wouldn’t suppose that the most convenient form for a presentation
of the properties of the language would be a dictionary. Rather, perhaps,
some form of sequence functions of a conversation, with various items
included in them. It would be hard, unless you knew ‘Hi’ was a greeting’ to
locate what ‘a greeting’ would be, via a dictionary. Whereas, if you had a
sequentially organized thing, if you knew what you needed you could see
where to find it. And you might also not be led to focus on the notion that
there are a set of alternative meanings for that word. That notion is an artifact
of an orientation to a dictionary-like object.

With data like this you can also see something which may be quite general,
and if it is of some generality it will be a helpful sort of fact in resolving
various puzzles. And that is, for the second introduction to come off as ‘an
introduction’, that there was a first, seen as ‘an introduction’ by its
participants, is sufficient to allow a rather different form to be used. “Ken,”
is all that’s used to signal that the second introductions have taken place. That
is to say, the item ‘second introduction,’ to be seen as having been done in the
way it was done, is contingent on there having been an observed ‘first
introduction. ’

If, given something that can come in sequences, a ‘first’ has been seen and
oriented to, then a ‘second’ can be done observably, without using that form
which would be required for a ‘first.’ That is to say, if the therapist had simply
said “Ken,” as a first, then what he was doing would not at all be clearly an
‘introduction.’

Here is the sort of thing that this phenomenon, if it is more or less general,
might clarify. It’s taken to be the case that there are various markings for
questions, e.g., rising intonation, a special grammar, inversion, etc. It has also
been suggested that an enormous number of observable questions have
neither ‘question intonation’ or the grammatical form of questions, and yet

288

Part 111

seem to be recognized as questions. How would that be done? One possibility
might be, if a ‘first question’ has been recognized, then things which go after
its answer can perhaps be inspected for whether they’re ‘second questions.’
And those can perhaps be made without use of those things which would
mark a ‘first question.’ In the data there are long sequences of things which
are treated as questions by the hearers of them, but where, increasingly into
a series, you just don’t get the markings which would in isolation show that
those were ‘questions.’ Just as you don’t get the markings for “Ken,” that
would show in isolation that it was an ‘introduction’ though “Jim, this is Al”
would be so seeable. And with respect to names in series, it’s a regular thing.
For example, if you’re offering candy, then you can perfectly well use this kind
of form: “Would you like a lifesaver Mary?,” “Bill?,” “Harry?” etc.

So then, first items in ‘ serialhables may have special status, and also, that
you are making ‘a series’ can then be done by doing as a second, something
which is not independently seeable as such a thing, but is so seeable given the
first.

One thing we can come to see is that producing the introductions in the
form of a sentence might specifically be done to make available to Jim that
‘a group’ is being presented. That is, we want to differentiate between Jim
being introduced to ‘three people’ in close order, and Jim being introduced to
‘the group,’ one by one. In that regard, it seems to follow that sentence-
making is to be conceived as a kind of social institution in perfectly
conventional ways. I suppose we don’t ordinarily think of the use of grammar
as a social institution for demonstrating organization. Courses in social
organization don’t have, I suppose, sections on the way you can build
sentences to present a group, where you use the resources of the grammar to
do that. But it might not be a bad idea.

It isn’t, then, that we have sentence-making on the one hand and social
structures on the other, and one can study their relationship by, e.g., studying
dialectics. But you can see that if we can propose that an introducer could go,
“Jim this is Al,” “Jim this is Ken,” “Jim this is Roger,” then the responses
of second and third persons would not show that they were attentive to what
had preceded, and furthermore, talking about them as ‘second’ and ‘third’
would have no specially strong basis. So the very sentence that’s used to make
the introductions does it specifically as ‘a round of introductions,’ involving all
the parties in seeing it as ‘a round of introductions’ in order for them to
perform its various parts.

We can also notice that, as in the case at hand, names used in introductions
seem to involve the doing of an instruction to the parties concerned, as to what
they may call each other. And that makes the matter initially interesting, in
that we could then ask, “How does A get off telling B, C, and D what they
should call each other?” Where it’s perfectly plain - though of course it will
have to be dealt with itself - that not anybody has any right to call anybody
else any particular name they may choose, of the various alternative ways that
names are used. I once asked a class to sit down and write out, over a day,
what versions of their name they were called. And they found, first of all,

Lecture 04.a

289

considerable orderliness to that, but by no means that they were called a
name. Nor did they figure that the various persons on their list could
interchangeably use the various names that were used by one or another
person. And not only could X not pick up the name by which Y called you,
but that you could not insist that X do so if you wanted to. And this is apart
from issues of the way in which what name they call you is or is not
informative about what name you should call them.

That the doing of naming here is also the doing of an instruction is most
interesting because it involves two different issues. One of them turns on how
you can go about doing instructions. The other is, is it generic that whenever
persons do ‘naming’ they’re also doing something else? Perhaps, indeed, that
they do ‘naming’ involves a procedural rule which says, When you find that
they’re doing “naming,” look, then, for something else that they’re also
doing.’ That would have some considerable consequences.

In that regard, let me just note that we’re getting here, not merely an
instruction, e.g., for Jim how to call Ken and for Ken how to call Jim, but
a presentation for all how to call each. Each usage is instruction not only with
regard to its participants, but also in some way to the others, where the use
of names in multiparty conversations is something which is to be considered
differently than the use of names in two-party conversations. That’s obvious
in certain ways. Persons modify the names they call each other in two-party
conversation when they are in a multi-party conversation. Furthermore, they
might refer to somebody in multi-party conversation differently than they
might in two-party conversation. Now in doing that they might perhaps be
taking account of the instructional use of names, and also, of course,
controlling the kinds of information that can be gotten about the two parties
concerned. That is, e.g., if you overhear two parties, and they use one or
another name to each other, you can make interesting guesses about their
relationships, and one of the ways they can intend to conceal that sort of thing
is by the kinds of names they use to each other ‘in public.’

I want to deal, now with a problem which we can call the ‘sufficiency of an
identification’ problem, the issue being: Are names sufficient identification, or
is it the case that, given a name, some other identification is thereby made
relevant? And it does seem that names provide for the relevance of other
identifications. In this two-hour group thereapy session there’s a couple of
places in which names are offered. For example, earlier the entrant’s name is
mentioned in a sequence that goes:

Al : Where’s Jim Reed?

Ther : Jim Reed will be late.

Roger : Who’s Jim Reed?

Al : New guy is coming.

I suppose if names were sufficient identification, an appropriate answer to
“Who’s Jim Reed?” would be, ‘Jim Reed is Jim Reed, who else could he be?’
Or, if the question had been formed up as ‘Who’s he?’ then the name could

290

Part III

be returned. In this case, the answer is, “New guy is coming.” At another
place somebody else is mentioned:

Ken : How has Fido across the street been treating you?

Al : Oh you mean Bitch Hazel?

Ken : Yeah heh

Roger : Who’s Fido?

Ken : ’s Hazel.

Al : Bitch Hazel across the street a waitress.

So, “new guy” and “a waitress.” And it is of course routinely the case that
in the beginning of first interactions, the exchange of names is accompanied
by the exchange of other identificatory items.

Let me make one caveat to the remark that further identifications are
appropriate: It is perfectly conceivable that although only names have been
used aloud, the very structure of interaction is sufficient to provide that some
other identifications obtain and are relevant, so that the exchange of names
will have informed those present about both the relevance of some other
identification, and the findings on those matters.

This can have to do with, for example, what it is in the first instance that
might be known about any one of the people involved, so as to use that to find
other things. So, one possibility is that what Jim, the entrant, knows is that
the guy who says “Hi Jim, come on in” is ‘the therapist,’ and that there are
also ‘patients’ present. The fact that he’s introduced to the three other guys
may be perfectly sufficient to tell him that they are, indeed, ‘patients. ’ And it
may be perfectly sufficient to tell them that he is, indeed, ‘the new guy.’ So
it’s then perfectly possible that the relevance of other identifications are given
to the various personnel present via, say, who greets who and who does
introductions, even though none of them are laid out directly. If that were so,
then the relevant identifications would vary with any given occasion of
greetings and introductions taking place, i.e., you wouldn’t say that any time
somebody introduces three boys to another, then the one doing the introduc-
ing is ‘the therapist’ and they are ‘patients.’

That being the case, we haven’t made a trivial remark because it would
have to be an issue that all of them decide what the import of that fellow
doing the introductions is, i.e., what he doesn’t have to say out loud. That
then makes the greeting-introduction sequence of possible significance to the
issue of other identifications than those explicitly used. And that has one
consequence methodologically which is, you couldn’t then say ‘You’re wrong
about the appropriateness of further identifications if there are some circum-
stances where they didn’t happen to occur.’ You might find a place where they
didn’t occur, where analysis of that case would show that the structure of the
introduction did that work. Though the question of how it did it might be
another issue.

One import of that fact, then, is its relevance for the question of what it
would take to show that a piece of analysis was wrong. And one thing that

Lecture 04.a

291

won’t do is some totally unanalyzed piece of interaction used to claim that
some possible rule, for example, was ‘wrong’ by virtue of that piece of
interaction not doing what the rule states. You must have been making
findings about that piece of interaction to propose that it controverts a
proposed finding.

Lecture 04. b

An introduction sequence (ctd)

I’ll continue picking away at this introduction sequence.

They.

Jim this is uh Al,

Jim:

Hi

Ther :

Ken,

Jim:

Hi

Ken :

Hi

Ther:

Roger.

Roger :

Hi

Jim:

Hi

Ther:

Jim Reed.

Among other things, I made an extensive — if not labored — consideration
of the fact that we could talk here of a ‘round of introductions’ having taken
place. And I proposed that there was machinery which could be adapted to
presenting, specifically, that a ‘round’ was being done. That is to say, the way
that that sentence, with its connections, comes off.

Now one might find oneself trying to figure out why it should matter
much that we could have a notion of a ‘round’ which specifically, distinctively
characterized some phenomenon; why, for example, we couldn’t talk of
whatever we pleased as a ‘round’ as long as, say, it might have gone to
completion, i.e., there were three people that could have been introduced, and
three people were introduced. Or in any event, if a somewhat picky notion of
‘round’ does characterize something, while that may be of incidental interest,
what will it permit us to come up with?

I am prepared to present some consequences of it, but before doing so, I
want to mention something that is not an altogether incidental sort of issue.

People often ask, ‘Why do you choose the particular data you choose? Is it
some problem that you have in mind that caused you to pick out this group
therapy session?’ And I’m very insistent that I just happened to have it,
somebody had found this segment, it became fascinating, and I spent some
time at it. Furthermore, it’s not that I attack it by virtue of some problem I
bring to it.

As we play around with what it turns out that we can get out of a notion
of ‘round,’ one thing to see is that it wouldn’t be supposed in the first instance
that, had we such a problem, we would have taken this data — with which we
come up with a solution for that problem - to consider that problem with in
the first place. But it may be, as it seems to be here, that we can be attentive

292

Lecture 04.b

293

to the possibility that some first-level analysis of a piece of data can set up
another consideration which will turn out to address, say, some theoretical or
methodological problems that we know exist in the field. And that is then to
say, there may be some real gains in trying to fit what you can hope to do, to
anything that happens to come up. I mean that not merely in the sense of
‘pick any data and you’ll find something,’ but ‘pick any data without any
problems in mind and you may find something.’ And how interesting what
you may come up with will be, is something you cannot in the first instance
say.

Now I adopt what is in a way a counter-strategy to the issue of
‘interesting,’ and that is that I’m specifically picking utterly uninteresting
data. Things which do not have for us any special lay interest. That means
that in order to find its interestingness we have to find that whatever it is that’s
interesting about it is what we can say about it. And we can then develop
criteria of interestingness where we’re not exploiting kinds of things we ‘want
to know about’ - scandalous topics, gossip, etc. Things like an exchange of
greetings are kind of ideal, rather than, say, the discourse of kings or salon
conversation, where we know in the case of the latter that it’s important and
interesting, and it’s very hard in the first instance to ignore ‘what they say,’
which you have to do.

Now, last time I talked about the introductions being brought off as a
‘round’ in a strong sense, i.e., discriminatively of saying ‘three people were
introduced here.’ Once we have that strong sense of the notion of a ‘round’
at hand, we can of course note that some sorts of things can come off in
rounds, others apparently are properly done only in rounds. For example, if
you have some amount of candy that you might offer, apparently you’d better
figure whether you have enough for any person present who might say ‘Yes,’
if you're going to make any offer at all.

Except that if that’s the way things somewhat work, then you can use the
phenomenon of a ‘round’ for some activities. For example, that offers ought
to come in rounds provides that you can specifically mistreat people. That is
to say, without ‘doing anything to them’ you can mistreat them by the not
doing of that thing which is underway in an observable round. For example,
Al, Ken and Roger are present, and if Jim offers some candy, “Would you
like some candy Al? . . . Ken?’’ and then it stops, he is observably ‘not
offering candy to Roger. ’ So there are specific things that can be done by virtue
of persons’ orientation to the fact that some activities are properly done in
rounds.

Now, leaving aside the issue of ‘mistreatment,’ when we consider that sort
of thing we can arrive at a decent way of talking about something that social
scientists always want to talk about, and it’s at least my reasonable guess that
they’ve never been able to talk about it in a strong way. And that is, how do
you go about talking about absences. Something is absent. Something didn’t
happen.

In pretty much any extended anthropological or sociological discussion
you’ll find them talking about something that didn’t happen. There’s a

294

Part III

tremendously tricky problem about talking about things that haven’t
happened, and that is, if you’re going to say some X hasn’t happened, then
there’s an indefinite list of things that could be said not to have happened at
that point. And if that’s so, then your case is non-discriminable from the rest
of those, and the observation is trivialized. ‘Sure he didn’t do X, but he didn’t
do A, B, C, etc., either.’ And how do you go about locating, as a significantly
sayable thing, that he didn’t do, distinctively, X?

The issue of arriving at a discriminative notion of an absence - something
that didn’t happen - is one that we should like to face. As Members in this
society we after all know that people are always saying things like “Mary
didn’t come’’ or “He didn’t hit me when I said this,” etc., but it is only by
knowing that you can do such things that you accept such statements as
reasonable. If you required that they be defended, they couldn’t be.

What we can perhaps do is see whether there are some methodical ways
that persons arrive at such noticings. In the first instance, they don’t go around
saying, e.g., when characterizing who was at a party, ’X was there and Y was
there,’ and then some list drawn from a telephone book to say who was
absent. Or, ‘He did X and he did Y’, and then some list drawn from a
dictionary to say what he didn’t do. There are some occasions under which
absences are noticed. If we can characterize the bases for them, we can come
up with a usable notion of ‘absence.’ And such a notion could perhaps be
generalized beyond the specific occasion that we happened to construct it in
relation to.

That is to say, if it’s the case that something appropriately goes in a round,
and what is observably a ‘round’ has begun but is not brought to completion,
then, that something was ‘not done’ is noticeable, and the specific thing not
done is noticeable; for example, that Roger was not offered a piece of candy,
or that Jim was not introduced to Roger. Then we can see how it is that
persons can come to talk non-trivially about absences.

Furthermore, it might be the case that what was being done with that thing
not happening is also observable. That is to say, it’s not merely the case that
somebody can apparently say that Roger wasn’t offered a piece of candy, but
that they can say what was being done with Roger not being offered a piece
of candy. There is, in that regard, material which consists of people remarking
on it e.g., ‘How come you didn’t offer me any?’ And such things might
permit us in some cases to get at what sorts of things come in rounds. So, for
example, in our data there’s a place where they’re telling jokes. A1 tells a joke.
Ken tells a joke. After a while Roger says “How come I never remember any
jokes,” i.e., he sees that he ‘hasn’t told a joke’ and remarks on it.

And jokes do come in rounds. Now, jokes and introductions and offers,
i.e., things that come in rounds, seem to have a feature in common with
something that’s partially related to rounds; that stuff I was talking about last
time, on firsts and seconds, where we could come up with a decent notion of
‘second’ independently of saying it happened to be number two, i.e., via
modifications in the way that a first thing got done. A first joke - what turns
out to be a ‘first joke,’ we could call it a single joke - may get introduced,

Lecture 04.b

295

“Hey I heard a joke. Wanna hear it?’’ Second ones routinely do not get
introduced that way, but by “Here’s a good one,” “Here’s another,” without
that it’s going to be a joke being specified. And that, then, is getting done in
the same way that, that this is a ‘question’ is not seen independently of the fact
that a series has been started, or that this is a ‘second introduction’ is not
available but by virtue of the fact that there was a first to which this has been
fitted. And it may be that non-trivial absence is not restricted to phenomena
that come in rounds, but that there are other ways that relevancies are set up
such that, that something didn’t happen is observable.

Once we come up with how, e.g., properly-rounded actions can permit us
to see ‘absences’ in controllable ways, we might be able to abstract from that
other ways that absences are, in parallel to our first cases, also observable. That
is to say, what’s going on here is that if it’s properly-rounded and a round has
begun - say, a first and second have been done, and done specifically as
rounded events — then a third is relevant and N is relevant until completion
has been arrived at. And what we’re dealing with, then, is ‘ contingently
relevant events. ’

It may be that you can formulate classes of contingently relevant events
independently of rounds, such that: Some prior thing has happened, and has
happened in such a way that you know it’s been observed as such, i.e. , in the
case of the ‘introduction,’ that an item was recognized as an ‘introduction’ was
provided for by the greeting sequence taking place, which turns on an analysis
of the ‘Jim, this is AT into an ‘introduction.’ You might find, then, that
wherever you get contingently relevant events becoming relevant, you can
find that absences are noticeable. And you can begin, then, to generate a fairly
abstract way to provide for distinctively using the notion of ‘absences.’

You may then find that a good deal of the sociological literature is not
‘wrong’ in the sense that it didn’t use absences that way, but that it is
specifically inexplicit and unknown. That is to say, it probably will often
conform to an appropriate use of absences, but it didn’t have what is essential,
and that is the apparatus that will give us that usage of ‘absences’ - where it
is our business to analyze how it is that something gets done, or how
something is ‘a something,’ and not to employ it.

Se we move from this nothing ‘introduction sequence’ to a notion of
‘rounds’ which handles a lot of things other than introduction sequences, and
then to a formulation of ‘absences’ which is, now, a perfectly conventional,
recognizable ‘problem.’ Everybody who’s doing social sciences wants to talk
about absences. And, that we extract some stuff on absences provides in the
end what’s interesting about the introduction sequence. Now we didn’t get it
by especially saying, ‘The next three lines of data I play with, I’m going to try
to deal with absences.’ And it’s not that we couldn’t have come up with a
decent way of talking about absences on other materials. But it might well be
other material which seems as irrelevant in the first instance to that issue, as
this stuff did.

Let me now address this ‘introduction sequence’ with a different sort of
question, concerning the fact of its placing, and what may be involved in that.

296

Part 111

It is the first piece of talk after the entry of the new member. Now the fact
that upon his entry this thing is placed, can lead us to considerations along the
following lines. They were, after all, talking about something or other before
he entered - automobiles, racing - and when this sequence closes, that topic
is again picked up. Can we say something like: Under some conditions there
are conversational possibilities which can be called priority items , or priority
sequences, and they have superceding relevance over whatever it is that may be
being done, or over some things that may be being done? It does appear that
there are some ‘priority items,’ and they’re priority items whose character is
not given by the last thing that occurred, in the sense that given a question an
answer may be a priority item, given a first greeting the second may be. But
these seem to have rights and obligations to be done, perhaps without regard
to what it is that has been taking place, under some proper conditions
occurring. They don’t have, in that sense, a ‘place’ in the conversation.

What I want here to point to are things like this: Suppose there’s a rule that
says: For some given person, upon the entry of some other specifiable person,
the first ought now to greet them and introduce them to others that they don’t
know, if they don’t know them. Now that’s an ‘instruction.’ There are
instructions of that sort written up in, e.g., etiquette books, and you may, in
fact, get instructed by someone about such a matter.

There are several interesting things about such instructions. First of all,
what are the ways that instructions directed to some determinable person out
of a group, are usable or properly to be used by others in controlling their own
activities? So that, for example, it may be the business of somebody, upon the
entry of Jim, to greet him and introduce him. However, if we had a situation
where the instruction had the character that it controls those to whom it’s
directed, then, for example, these fellows having been talking, they could go
on talking. And treat, for example, the fact that the therapist starts to talk to
Jim, as an ‘interruption.’ Now one of the characters of priority items is that
when they are properly invoked they are not interrupting what goes on, and
what goes on ought to cease. So that, for example, even if the instruction
about greeting and introduction isn’t directed to some set of people, they, in
the first instance, apparently ought to orient to that instruction so as to allow
it to get done. For example, to shut up, and perhaps in some ways to
participate.

If the character of at least some instructions is of that sort, then the task,
say, of sociologically rewriting instruction manuals could get at least an initial
posing. That is to say, if you have an instruction manual that has a set of
directed instructions, i.e., say, ‘A host, upon the entry of invited guests, ought
to do the following,’ then those instructions are not secrets from other
Members, but - crucial to their use - involve others seeing them and going
through some analysis in which they treat the instructions as instructions for
them also.

Now that doesn’t mean that one can just take a set of instructions, where
each instruction had a direction (to ‘host,’ to X, to Y) and now amalgamate
all the persons and say, ‘These are to whomsoever, everybody ought to notice

Lecture 04.b

297

them and attend them in one or another way. ’ But presumably each of them
has to be somewhat rewritten for each possible person present. The character
of the rewriting involves at core, among other things, some means whereby
upon, say, any priority item possibly being relevant, every person present has
some way of engaging in a systematic categorization of the various personnel
present, of locating among them ‘the one who ought to do the priority item,’
of doing what will allow that, e.g., being quiet or talking - indeed, of seeing
that such an item is present so as to stop in the first place. Where we need
some generalized results - unless we’re going to get difficulties, i.e., that each
person arrives at more or less the same determination of whose business it is
to do whatever it is that’s to be done.

And there’s some curious aspects to the issue of who’s business it is. The
instructions in etiquette books tend to get written with some specified type of
person, given some type of situation, e.g., ‘The host at a party ought
There are presumably lots of occasions in which, say, introductions occur, for
which there isn’t a ‘host at the dinner party.’ The question then is, how is it
that persons go about picking some class for which the set of integrated rules
has been initially written, and replacing the various categories with some
available set of categories, and assigning the new categories status with respect
to the old, i.e., seeing who is ‘like a host’ or some such thing. Or treating ‘the
host’ as an example from some other class that has some other form, i.e., ‘the
person in charge here,’ or whatever. Where that person can find himself, and
find the others seeing that he is the one, although he isn’t ‘a host.’ Maybe.
And all this is to be done in fairly short order.

So these ‘priority items’ apparently involve, then, some orientation to a set
of identifications, where whatever the instructions are, they’re usable by other
than the one(s) to whom they’re directed. And the results of whatever analyses
Members go through to find who does this, are apparently not controlled by
the fact that the instructions which weren’t in the first instance directed to
them, didn’t in the first instance specify some person who happens to be there.
That is, say there’s no host here, but people can still find some way of locating
that they are - and others see that they are - the one who should do this. And
they can organize their activities so as to allow that, and to make it missing
if it doesn’t occur, and various other things.

Now notice one extremely important aspect of that. In the case of greetings
and introductions, the analyses which provide for that set of concerting
activities had to be made doable ‘even though’ - if I may use that phrase -
the set of persons present don’t all know each other. They are not doing it, in
many cases, by virtue of having made some predetermination, or knowing
each other’s habits. In the case of introductions - and that’s one of the
interests in introductions - the whole apparatus has to be set up to be usable
between persons who haven’t met.

Consider the kind of import it has for some set of concerted activities,
where it’s crucial that they go off concerted, that there’s a bunch of analyses
that have to be done, which can be relied on, where the people who are
generating the activities don’t know each other in the first place. It is that such

298

Part 111

things go on routinely that is the sort of thing that sociology in the first place
picks up and focusses on - or ought to - when it talks about ‘a stable set of
social affairs’ or whatever.

You can have, apparently, stability and orderliness where people don’t
know each other, and are doing things of an order that you could see easily
enough if you tried merely to rewrite the instructions out of Emily Post to
handle the five people in this group. And all that can be done ‘immediately.’
Time and time again. And that is to say, these are ordered events on single
occasions. It isn’t the stability which sociology is considerably interested in,
which is gotten via summation of frequencies of occurrence. That isn’t its
beauty. That isn’t what Hobbes, say, might have been interested in. It’s that
sort of thing that provides why one ought to be fascinated by a thing like this
running off in five seconds - if five seconds is what it takes; it may take less.
You’re really seeing stuff with tremendous power at work.

Then there’s this matter of ’priorities,’ i.e., the kinds of analyses involved
in locating the fact that such a priority event is now appropriate — where, of
course, if one locates it and others don’t, then the one who locates it may be
heard to be merely interrupting, if other things are going on. And it isn’t the
case that it’s simply a rule that says ‘If any given person comes into a room,
any room, and doesn’t know all of the people present, they ought to be
introduced to them.’ If I’m sitting in my office talking to somebody, someone
will frequently come into the room and say something to me, and I don’t
bother introducing them to the other person present, who may pause, talk
after a while, various other things. We don’t have a universal rule, which
would make everything quite simple, i.e., a rule stating ‘Any entry to a room
immediately provides that this thing should take place.’ Then you’d need no
analysis.

It’s a curious fact that situations in which something holds for ‘everybody’
are far simpler than those which don’t have that character, but which hold ‘for
men,’ ‘for women,’ or whatever. Because then people apparently have to
locate what set of categories are appropriate, i.e., if someone enters who could
be a ‘man’ or a ‘doctor,’ or something else, then the Members have to know
that the device that contains ‘men’ is the one to be used. As long as it’s not
‘everybody’ there’s an issue of which device to use - and in the case of
introductions, it’s apparently a variety of different things; somethimes ‘age,’
sometimes ‘sex,’ or ‘proper being there,’ etc.

Again then: I made some remarks about ‘instructions’ which can have,
perhaps, some tremendous generalizability in terms of at least the problem of
dealing with instructions in general. One could ask on what basis could such
findings have been made? You could imagine that in principle what one
would have had to do to come up with them would be to consider a sample
of instructions; for example, ‘We have to take all the instruction manuals
available in some game book, and analyze some common properties of them,
see who they’re directed to, etc. ’ Alternatively, it may be that we can come up
with findings of some considerable generality by looking at very singular,
particular things. By asking what it takes for those things to have come off.

Lecture 04.b

299

Or, for example, would the sort of things that I’ve been saying so far be
affected by a study which examined 100,000 random greeting sequences, and
computed the sequences? Indeed, the point that one wants to notice is that one
couldn’t have done that, since by and large any such study would already
have come to a definitive formulation of who was present, in some set of
categories - ‘male-female,’ ‘age,’ ‘social class,’ etc., and those may be precisely
the things at issue in this thing coming off. That is, that’s what people have
to find out in doing them.

Lecture 4

Invitations', Identifications;
Category-bound activities

Ther:

Jim, this is uh Al,

Jim :

Hi

Ther :

Ken,

Jim.-.

Hi

Ken :

Hi

Ther :

Roger.

Roger :

Hi

Jim:

Hi

Ther :

Jim Reed.

( ):

((cough))

Ken: :

We were in an automobile discussion

Roger :

discussing the psychological motives for

( ):

hhhhhhh

At:

drag racing on the street.

The line I’m going to talk about this time is “We were in an automobile
discussion.” I will by no means have fully characterized it; that will take a
long time and I’ll keep coming back to it. My considerations here will be
another instance of how it is I take it that we go about doing sociology and
anthropology.

I want to propose that this utterance, “We were in an automobile
discussion,” at least were it to occur alone, is possibly to be seen as an
invitation. I want to lay out the tasks that are involved in showing that that’s
so, and then to attempt to satisfy them.

We want to be able to build a method which will provide for some
utterance as a ‘recognizable invitation.’ Two tasks - at least - are involved,
and these two are closely related. One task is to construct what I’ll call a
‘partial definition of an invitation.’ What makes it partial is that, while it’s a
way of doing invitations, it is clearly not all the ways. There are other ways,
and those would be other partial definitions.

Now, we can have a perfectly good partial definition of an invitation
which, however, wouldn’t give us this thing as ‘an invitation.’ The second
task, then, is to build a partial definition which provides for this to be a case
of the use of that procedure.

That means, for the first task, that I could say, “It looks to me like an
invitation, and I bet it looks to you like an invitation,” but I wouldn’t be able

300

Lecture 4

301

to show how that’s so. If that’s the case, I don’t establish my point. For the
second task, I could also say, “Here’s the way some invitations are done,” and
we could find cases that would fall under that, but not this one. That, again,
wouldn’t do what I propose to do. We want to do both: Construct a partial
definition of invitation,’ and one that provides for ‘this is a case.’

And in the first place, we want a method that generates this line. We want
it, however, to do the job of generating it as a case of an invitation. It’s
perfectly easy, and perfectly trivial, to construct methods that would generate
this. We could use the spelling rules of English. We could say, “Divide the
thing in half. First put the first half in, then put the second half in.” That
would not, however, tell us how whatever we get is recognizable as a
particular activity - like ‘invitation.’ We want, that is, to construct a
discriminative procedure for generating invitations.

Further, since invitations stand in alternation to rejection, what we would
like is to discriminate between the two. We want the definition to provide for
the recognizability of ‘invitation’ for some cases, and for the recognizability of
‘rejection’ for some. And if we get a method, then we ought to be able to use
it to generate other cases than this one; where, then, the ones that we generate
ought to be equally recognizable as invitations or rejections. With that, we
have some idea of the tasks involved.

I’m going to focus, basically on the second part, the phrase “an automobile
discussion,” as crucial for our purposes here. Now, I’ve introduced the term
‘category-bound activities.’ I want to propose a relevance rule, and that is: If
someone names - as, for example, what they were doing - a category-bound
activity, that provides, first, for the relevance of the category to which the
activity is bound. Then, by use of the consistency rule, the collection of which
the category is a member is made relevant. So that by naming as the thing you
were doing, some activity which is category bound, you provide the relevance
of some collection of which it’s at least the case that some particular activity
is bound to a category of that collection.

And an automobile discussion is, I think, a category-bound activity, with
‘teenage boys’ as the category it’s bound to. By naming the activity, one thing
that we can say is being done is that an identification is being offered, of at
least the speaker. That is, that with this formulation he says, “It is relevant
that we are teenagers; consider us teenagers.”

Essentially what I want to propose is that one way to make invitations is
to do the following: If a new person comes into some place where there is
some set of persons engaged in, for example, talk, then one of the pre-present
persons can make an invitation to the entrant by using a phrase to start off
with, like “We were doing,” plus the naming of some activity that is
something like ‘category-bound’ to a category which, once made relevant -
and it’s made relevant by the naming of the activity - can be seen
convergently to hold for some part of the pre-present persons and the entrant.
Alternatively, one can build a ‘rejection’ in the following way: Propose “We
were doing” plus a category-bound activity, where the activity is bound to
some category that characterizes the pre-present persons, but has the entrant

302

Part III

in an alternative category of that collection. For example, a man comes home.
His wife is there, talking to a friend of hers. She says, “We’re discussing the
sewing circle.’’ It’s a rejection. He says “Oh excuse me.” That is to say, that
thing is bound to ‘adult females’ and he’s a male. It provides for them as
members, and not for him.

Now, it’s perfectly clear that we could find this statement, “We were in an
automobile discussion,” when there wouldn’t be an ‘invitation.’ For example,
you get home, somebody says to you, “What did you do?” and you say, “We
were in an automobile discussion.” We need, then, a slot in conversation
wherein items, if they occur there, can be recognized as invitations, if they are
in the first place possible invitations. I want to propose that that point in a
conversation after greetings have been exchanged, and at least a partial
introduction, is a slot for an invitation. For one, that means that one can look
to that place to see if something that passes as ‘an invitation’ has occurred
there.

There are, in the ‘invitation slot,’ alternative but related things that can also
go, and it’s rather relevant to see that. Among them are things which we
could call ‘ pre-invitation Injections (with this term I intend that the thing can
be either pre-invitational or pre-rejectional). The matter of pre-invitation/
rejections and their use can be rather consequential, and I’ll therefore quote a
piece of data and its analysis from one of the extremely few studies of
conversation that have yet been done; a book called The First Five Minutes ,
by Pittinger, Hockett, and Danehy. The quotation occurs right at the
beginning of this five-minute segment of a psychotherapy session which they
are engaged in analyzing.

Therapist: Will you sit there,

(apparently she sits there)

Therapist : What brings you here?

Patient : Everything’s wrong. I get so irritable, tense, depressed.

Just everything and everybody gets on my nerves.

Therapist : Yeah.

Patient : I don’t feel like talking.

The key line is “What brings you here?,” and this is what they say about
it:

This is T’s opening gambit. It would seem about as open-ended and
non-directive as it could be made, and in many ways it is, as can be
shown by comparing it with other possible openings, but in one respect
it tips the scales slightly. Three of the four words in the sentence; what,’
‘you,’ ‘here,’ are what the grammarian calls ‘substitutes’ or ‘shifters’ or
‘pronominal forms,’ words the denotation of which in a specific context
depends almost wholly on that context. Instead of merely ‘what,’ T

Lecture 4

303

could have said ‘what troubles’ or ‘what problems’ or ‘what difficulties.’
The actual choice is obviously neutral. Instead of ‘here,’ he might have
said ‘to a psychiatrist’ or to this clinic’ or the like. The neutral ‘here’
allows P to define its reference in accordance with their own perceptions
of the surroundings.

That’s a very casual treatment, and I want to argue that it’s all wrong, and
the fact that it’s wrong operates to prohibit them from seeing how the patient
then comes to respond as she does, i.e., to say, “I don’t feel like talking. ” That
is, the doctor says, “What brings you here?” The authors say it’s a very open
beginning. I want to propose that it’s a pre-invitational/rejectional remark,
and quite obviously so. Things like “what brings you here” regularly occur.
You walk up to somebody’s office, walk in, say “Can I see you for a
moment?,” they say “What do you want?” That’s not an invitation. You
know at that point that now you say something, and they may then use it to
decide whether they’ll see you or not.

In a first therapy session, the troublesomeness of ‘What brings you here’ is
much larger than the chance of losing whatever it is the interchange was to
hold (that is, if the patient takes it that what she says next is going to stand
as some sort of grounds under some analysis of it, for the therapist to decide
whether he’s going to take her or not as a patient, whether he’s going to
continue beyond this initial diagnostic interview), but, further, what it is that
would be good grounds for him to take her is something she may not have
any idea about at all. It’s not enough that she’s in one way or another troubled
- that’s obvious enough in that she’s here. She could present a set of troubles
which he could recognize as legitimately complainable about, and something
to be cured - but perhaps not by him, or not now, or not in an out-patient
status, maybe in a hospital. What it is that he’s going to do, then, with an
item she presents, is something that she, it’s reasonable to suppose, doesn’t
know. And furthermore, if he turns her down after she’s told him something,
then she’s perhaps given somebody information which they aren’t entitled to
have. That is, if he’s her doctor, then her private troubles are something that’s
his business. If he’s not, they’re not. And if whether he’s going to be her
doctor is conditioned on what he makes of what she says, then, what she can
tell him which can also control that she hasn’t told him anything terrible if he
turns her down, is perhaps a rather problematic set of materials.

Now, she accepts the pre-invitation/rejection by offering what could be
good grounds for being accepted, and what she may take to be a sufficient
answer. That is, answers to such questions as “What brings you here?” are
regularly seen by the doer to be sufficient when finished, and, regularly,
patients do, if they figure they can’t deliver in an utterance what they take to
be sufficient, say things like “Well, it’s a long story,” to provide that they can
then at least go through that ‘story’ in its particulars. I take it, then that in this
case she gives what she figures to be a sufficient answer. And it’s rejected. That
is to say, the doctor’s statement, “Yeah,” would then be heard as saying
things like “Continue,” “What else?,” i.e., ‘That isn’t enough.’

304

Part III

Note what I’ve proposed about “Yeah” and its possible equivalents, and
compare it with what is being proposed as possible alternatives to the
opening “What brings you here?” (e.g., “What troubles bring you to this
clinic?”). One would like to ask how is it that they go about deciding what
the population of possible openings is. Why is it the case that something
altogether different from “What brings you here?” is not to be considered
as a possible alternative to it? One possible reason might be that, having
made no systematic formulation of its properties, they have no way, except
in an altogether lay fashion, of providing what are equivalents to it and
what are variations on it. And then we get what is a perfectly naively
obvious way of talking about variations: Drop one word out, stick another
word in.

I want also to note that the status of “What brings you here?” as a
pre-in vitational/rejectional needn’t be dug out of therapy conversations at all.
It, and things like it, occur in a great mass of non-therapy situations. One
could well have formulated a characterization of them, seeing their status as
pre-invitational/rejectional, without ever looking at therapy. And that, of
course, is what would provide for the patient, not having been in therapy
before, seeing what it is that’s being done. What I’m concerned with here is
to provide that persons will be detecting what various items are doing, and
handling them, via the fact that they’ve dealt with others.

Returning, now, to this slot in which we’ve found ‘invitations’ (and
rejections) and ‘pre-invitational /rejections,’ another thing that goes here are
pure ‘pre-invitationals.’ They are things like, for example, in a phone
conversation:

A: Hello

B: Hello

A: What are you doing?

B: Nothing

In that slot, “What are you doing?” seems clearly to be pre-in vitadonal. The
fact that it is - and is seen to be - so, gives us some sense to a clarity of a
certain term which, if you figured yourself to be a tough-minded sort, you
might say was characteristically unclear; that is, “Nothing.” Clearly enough
he’s doing something; talking on the phone, breathing, and the like. It’s given
the character of the first as pre-invitational that “Nothing” is clear. That is,
it accepts the pre-invitational status of the question and now awaits the
invitation, whatever it may be. And I’ll consider items like “Nothing” in
much more detail later.

The fact that there are pre-invitational items alone warrants some
consideration. It may turn on things like the following: Rejections are
countable things. That is to say, you’re supposed to use them so as to decide
whether to generate another invitation, and you ought to learn from a
rejection whether another invitation ought to be made. On the other hand,
someone can take it that if they’ve engaged in some sufficient amount of

Lecture 4

305

rejections, the inviter will give up. If that’s so, then there may be some
attempt to use items which are not subject to starting a count, such that what
you do is propose that you might make an invitation if an invitation will be
well looked upon. If there is some indication to the contrary, you don’t make
the invitation, and thereby you get no rejection. And thereby, perhaps, you
have the right to start over again without having a score.

Now, one thing I want strongly to emphasize is the way that our analysis
will give us constraints on what it is that’s equivalent, and what are serious
alternatives to that which occurred. So, for example, one thing you would
want as an alternative would be something that’s a case of another class;
where the classes stand in alternation. And in that regard, we could construct
alternatives to “What brings you here?’’ which were alternatives by virtue of
not being pre-invitation/rejections, but which were pre-invitations, or invi-
tations, or other sorts of things which can also go in that slot.

Lecture 5

Proffering identifications ; The Navy
pilot ; Slots ; Paired objects; Adequate
complete utterances

I’ll continue our consideration of this sentence, “We were in an automobile
discussion.” One of the things I’ll regularly do when I propose that some
activity is being done in some fashion, is to provide alternative independent
materials which show, I take it, that the proposed activity does get done, and
in such a fashion. That is to say, if I sometimes say, for example, that one can
make an identification by asserting some membership category as one that one
would belong to, then it may not be obvious that such a thing is ever done
at all. It may appear that I just made it up as a way of talking about this
particular thing I’m trying to talk about.

So, then, for the proposal that one makes identifications by asserting some
membership category as one that one would belong to, I want to note that:
That membership categories have that character is something which is quite
obviously not restricted to this society but is rather widespread. Here is just
one of what I suppose could be an endless string of quotations to that effect.
This is from a paper called ‘Social and psychological aspects of education in
Taieland,’ Africa, 11, 4 (1958), suppl. p. 46, by Meyer Fords. He writes:

If one asks a child “who are you?” the answer is invariably “I’m

so-and-so’s child,” or “of so-and-so’s family.”

The answer to “Who are you?” then, being the naming of some membership
categorization device (family) or category (child), etc.

One further thing I proposed, which I want to pause to show can be done,
also concerns the phenomenon of an identification. I have used, and I didn’t
intend it as a casual use, instead of saying that one ‘makes an identification,’
I’ve said ‘one proffers an identification of oneself.’ Those are different things,
where, for proffering identifications, it’s quite irrelevant whether you’re
correct, for example, but the issue is, again, whether it’s relevant that you’re
correct, or whether some other identification is, or is assumed to be, more
relevant than the one you propose.

There are circumstances where what’s central about a proffered identifica-
tion is the following: That there is some population besides yourself to be
identified, and that if they accept the identification you proffer about yourself,
you may have a very good idea about what identification they use for

306

Lecture 5

307

themselves, i.e., for one, that what they will use is an identification from the
same device; one that is consistent with the one you’ve used. And in the case
at hand, that will be quite crucial.

Here is a rather striking instance of the same thing, in which can clearly be
seen the proffering of an identification for oneself and for others, as well. It
comes from the New York Times , Saturday May 29, 1965. It’s an interview
with a navy pilot. The headline is: ‘A Navy Pilot Calls Vietnam Duty Peak
of Career.’

How did he feel about knowing that with all the care he took in aiming
only at military targets, someone was probably being killed by his
bombs?

“I certainly don’t like the idea that I might be killing anybody, but
I don’t lose any sleep over it. You have to be impersonal in this business.
Over North Vietnam I condition myself to think that I’m a military
man being shot at by another military man like myself.”

What he’s doing there, for one thing, is offering to those people who are
shooting at him, that he will consider them ‘military men’ if they will consider
him a ‘military man.’ And that if they both do it, then there are well-defined
rules for playing that game.

I say ‘‘playing that game;” I don’t mean it ironically. What I mean is that
the most characteristic places in which this proffering of identification within
a consistent set of categories is done, is in games. There it is quite striking that
if you say “I’ll be third base,” unless somebody else says “and I’ll be. . .”
another position, and the others say they’ll be the other positions, then you’re
not that thing. You can’t play.

Now, in the case I’ve just quoted, my remark about the proffering of an
identification is rather relevant, in that it happens to be the case — and these
guys know it - that the North Vietnamese don’t want to use that device
which provides for military men versus military men.’ They want to propose
no ‘war,’ no ‘military men;’ these guys are criminals. They want, then, to call
him a ‘criminal,’ and treat themselves as ‘doing police action,’ for example.
And there are rather different consequences, of course. For one, things like the
Geneva rules don’t apply, etc. The situation then is altogether different
depending on which is used, and whether the one he proposes for himself -
and thereby the one he proposes for them - is accepted.

So, again we can see that something which we can call ‘proffering an
identification’ is done, and is relevant, and therefore it’s at least possible that
such a thing could be being done in our materials. Whether it is, is another
question, of course.

What I want to do now is return to another partially loose end of prior
consideration, and consider what I’ve been talking about as ‘slots,’ and from
there, move to further considerations of structure. I said that one of the ways
that “We were in an automobile discussion” could be recognized as an
‘invitation’ was by virtue of the fact that it occurred in a slot where such

308

Part 111

things properly occur. Then I gave a list of other things which can also occur
in the same slot: ‘pre-invitations’ and ‘pre-invitation/rejections.’ Now let’s
consider this notion of a ‘slot.’ Are there such things?

What’s involved, for one, is a notion that there is something we could say
to be a natural unit, which you could call a ‘conversation.’ It’s not obvious
how we would go about showing that there is such a thing, or how we would
go about dealing with the way it’s composed. One way is to propose that it
has some invariable parts such that they always occur in anything you would
want to call a ‘conversation,’ and when they don’t occur, you don’t have a
‘conversation.’ Unfortunately, I don’t think there are such things. What’s an
alternative? An alternative would be, perhaps, that there are invariably
relevant parts such that if they occur you have a ‘conversation,’ and if they
don’t occur, you could say they are ‘absent.’ That sounds very queer but I’ll
talk about ways in which we can come to be able to say non-trivially about
something, that it didn’t occur. I say ‘non-trivially’ because first of all, you can
trivially say, when anything occurs, that an indefinite number of things didn’t
occur. The question of being able to say that some particular thing is absent,
didn’t occur, involves having to say that it is expectable that it did; that there’s
a place for it so that you can look to see where it was missing.

Assuming that we would have a unit, ‘the conversation,’ there is an object
that is ‘ahistorically relevant’ for conversation (i.e., there is no rule of
exclusion, such that it can occur in any given conversation). That is, of course,
‘greetings.’ One thing that’s nice about greetings is that the greeting items,
things like “Hello,” are not greetings wherever they happen to occur. They
are only greetings if they occur in the greeting place. And that’s a pretty good
way to see that we have a greeting place. And the greeting place is what we
could call ‘first position.’ If the item occurs there, then it’s a ‘greeting.’ We can
also look there to see whether it’s absent, because we know where to look.
Since we have no rule of exclusion for it, we have a right to look to see
whether it’s in ‘first position.’ And if it’s not, we have a right to say it’s
‘absent.’

So we do have an item which gives us some basis for saying there’s
something which we can call a ‘conversation.’ It not only has some things that
belong in it, that in that sense define it, but it also has at least partially, a
sequential structure, i.e., there’s at least a ‘first position’ in it which has some
item that belongs in it.

Now we want to dig into ‘conversation’ to see whether it has a deeper
sequential structure than that; for example, how far into it there is a general
sequential structure, and if not, what other sorts of sequencing rules are there.
What I think we’ll find eventually is that for conversation generally there’s a
very shallow sequential structure in the sense of general slots and items, but
that there are a great many other sequencing rules which are operative.

Among the other sequencing rules are things of the following sort - what
I’ll talk of as ‘ paired objects such that if a first member of a pair occurs, then
the second ought to be done, and if it’s not, that’s noticeable. The simplest
case of such paired objects is, again, ‘greetings.’

Lecture 3

309

In the case of ‘greetings’ their simplicity resides, for one, in the fact that if
A offers a case of the class, then B properly returns with the same case of the
class or another case of the class. Now, for some of these paired objects the
classes are two different ones, i.e., if A offers a question then B comes back
with an answer.

The question-answer pair is not embedded in the overall structure of a
conversation, which would mean, e.g., that it could only occur in third
position. It can occur in many places, and can run on indefinitely, as I
mentioned earlier, by virtue of this ‘chaining rule’ (that a person who asks the
question has a right to talk again after the other talks, and has a right to ask
a question). We see immediately that if, after ‘greetings’ whoever it is that has
the chance to talk asks a question, we already have a conversation, not only
as a ‘natural unit’ but we have it as a unit that is in principle infinite in length
(under the usual definition of ‘infinite’ - every conversation is, of course,
finite - ‘infinite’ meaning whatever the length, it could always be an item
longer).

Let me just note that I’m making no proposal like ‘one utterance, one
activity’ or anything like that. There are lots of things that can be combined
addatively, like “Hi, how are you feeling?,” or they can be combined by the
same item doing two different jobs - as in the case at hand I’m proposing that
‘invitation’ and ‘identification’ are being done in the same item, “We were in
an automobile discussion.”

By the way, items that have these positions, if for one reason or another
haven’t been used in their positions, can get used elsewhere, recognizably,
when they’re begun with things like “by the way.” For example, suppose you
don’t know somebody, such that if you were somehow brought to make their
acquaintance you might exchange names in ‘first position’ with them, as that’s
where ‘introductions’ tend to go. However, if you happen to have asked them
a question which doesn’t involve your presenting yourself as though you’re
about to get into a conversation or make an acquaintance with them, then if
it turns out that you do make an acquaintance with them, get into some
conversation, you can at some point say, “By the way, my name is
such-and-such.” Or, for example, you feel you’ve made a breach of etiquette:
An acquaintance comes up to you, they say “Hi, how are you?” and now you
go into a 15 -minute conversation in which only you talk and they say “yeah,
yeah,” and you just talk about your troubles. They finally end the conversa-
tion and start to leave, and you might say, “By the way, how are you?” as
they’re walking away. So the use of ‘by the way’ is a way of seeing that the
items that it is used with have proper positions and this is not in its position,
where ‘by the way’ identifies it as the thing it would be in its proper position.

Now, one of the things we’re getting to in this consideration of structure,
concerns the following. Two-party conversations are rather different things
than what I’ll call ‘multi-party’ conversations (‘multi’ always meaning ‘more
than two’). Because two-party conversations have an almost invariable
structure, and that is: Two-party conversations go A-B-A-B, etc. This may
sound like an immensely trivial observation. The first way to see it as

310

Part 111

non-trivial is to see that three-party conversations don’t go A-B-C-A-B-C
indefinitely, and four-party conversations don’t go A-B-C-D-A-B-C-D indef-
initely, where you might say it would be trivial, if those others held, that
A-B-A-B was asserted. Further, what makes A-B-A-B non-trivial is the fact
that I’m not proposing that ‘A’ is any utterance of the first speaker and ‘B’
is any utterance of the second speaker, etc., but that we can, in fact, come up
with an object which will, at least in part, be operative, which we can call an
‘adequate complete utterance ,’ which is not necessarily at all a sentence, and
which is a proper object in A-B-A-B.

An ‘adequate complete utterance’ is, roughly, such a thing as can be
recognized, when it’s done, to be finished so that if the other person ought to
speak, they can see when that’s done, and speak. The notion that it can be
recognized when it’s completed, and thereby can be produced to be
recognized as complete when done, has some kind of importance.

For one, let’s say there’s a kind of behavioral definition of a sentence which
approaches this; for example, Whorf s. Fie talks of sentences being some
grammatically coherent string, at the end of which there’s a pause. Others also
talk about the pause as characteristic of an utterance. But this cannot be
supported, since a pause is an utterly equivocal object. The question is, is it
your pause or my pause? Have you not yet finished and you’re pausing, or am
I supposed to be talking and therefore I’m pausing? And the way we get at
that has to do with the question of whether there was, for example, an
‘adequate complete utterance’ yet. I am not, however, proposing that it’s
necessary to stop at the end of an ‘adequate complete utterance.’

One relevance of ‘adequate complete utterances’ is that we could then
formulate a conversation which partially went, not A-B, but A-A-B, for
example; which, if we had as our formulation that the first person talks until
he stops and the other person talks, we could never come up with. And there’s
very good reason to try to do that, since people regularly operate on things
like: “I’ve already talked enough,” “I’ve finished what I have to say, why
aren’t you talking?,” “and now I’ll start up again,” etc.

Now the simplest kind of members of the class ‘adequate complete
utterance’ are, again, ‘greetings.’ “Hello” is an adequate complete utterance.
It is also sequentially relevant. At its completion, second speaker should pick
an item from the same class, and speak also. And it’s recognized when
finished. There are other items which serve as adequate complete utterances;
for example, questions are regularly that sort of item. When I’ve asked a
question, the pause between my talk and yours is your silence.

That’s a rather relevant matter. Suppose you’re engaged in questioning a
witness in a law court, congressional hearing, or your spouse. And what you
want to show is that something you figure they’re not going to tell you is
something they ‘know the answer to.’ One way you can go about it is, you ask
them a series of very routine questions. When you do that you get a
‘normative pause’ on their part. “What day is it?,” “Saturday,” etc., etc. At
some point, then, you introduce your question and watch the pause. The size
of the pause, and their own response to the fact that they see that they’re now

Lecture 5

311

‘avoiding,’ is sometimes perfectly obvious, e.g., they’ll turn red as they see the
length of the pause go, and it’s perfectly clear to whoever is watching that
they’re not telling something, if there’s something it’s been proposed they
know. That could only be, of course, if it were the case that one could assign
whose the pause was.

Now, questions are not only adequate complete utterances, but they are
‘ utterance completers in the sense that if you’re talking along for some length
of time, then one of the ways you bring your own utterance to ‘completion’
and provide that the other ought to talk, is by appending a question. The way
to get participation in a monologue is by using your monologue and simply
inserting, in various places in it, question intonation, providing only for “uh
huh,” ‘‘yes,’’ etc., without pausing any more than that. For example, at the
beginning of this group therapy session we get:

Ken : Hey- hey check this I went down to Hollywood Automotives?

Al: Mm hm,

Ken : And they told me . . .etc.

That way you get a conversation that runs A-B-A-B in which, let’s say, all the
Bs are ‘uh huh’s. And nonetheless it’s a perfectly normal and legitimate
conversation. It isn’t the case that any conversation requires such things as ‘a
balancing of talk.’

Lecture 6

Omni-relevant devices ; Cover
identifications

By this time it ought to be quite obvious that the correctness of a statement
does not of itself provide for its relevance and does not of itself provide for its
occurrence. So it would not be adequate to say that “We were in an
automobile discussion” was said by virtue of the fact that it is correct. Were
that so we could substitute an indefinite set of sentences, and the first remark
after the greetings had been exchanged could perfectly well have been, “Ice
floats on water.”

That may seem merely obvious. But the way it can cut is, if you read
transcripts you’re likely to read them through. And if somebody says things
that you agree with and you figure they’re so, you figure there’s nothing much
to say about it. And if you were doing conventional sociology all you would
do with them is put them into a table. Now, it isn’t that those things are any
less analyzable than anything else. The fact that they are or are not in an
automobile discussion in the first place, doesn’t resolve for you what your task
of seeing how come they said that thing and what they could be doing with
it, is. Going over the thing, seeing them saying things that you agree with or
things you figure are true, it isn’t enough to say “he said it because it’s true,”
or “there’s nothing there to study.” If he says something outlandish, that’s
what is seen to be studied. Of course the whole character of an education in
sociology, which supposes that what’s interesting to explain is ‘deviance,’
‘failure of integration,’ ‘role strains’ and the like, leads you to such a view.

This time, I want to say some more things relevant to our consideration of
“We were in an automobile discussion” as an invitation. I noted that while
there is a correct format for two-party conversations. A-B-A-B, I don’t know
that there’s such a format for multi-party conversations. That has a
consequence for us, in that whereas for two-party conversations if we know
who is speaking now we know who might speak next - and the party would
know the same - for multi-party conversations that would not be so.

Consequences of that for the doing of invitations is that at the end of the
exchange of greetings, who it is that ought to speak is not given via a formula
generic to that type of conversation, i.e., multi-party conversation, or a specific
n people in the conversation, or anything like it. Therefore, the behavior of the
entrant into the conversation is quite relevant to the occurrence or not of an
invitation in that slot. If he wants to see whether he’s going to get an
invitation or a rejection, then one thing he has to do is not fill that slot with
his own talk.

312

Lecture 6

313

However, whereas in two-party conversation it could usually be proposed
that a speaker is addressing the other party to the conversation, in multi-party
conversation a speaker may be addressing all or only some of those other
persons present. That fact will have some independent relevance for us as we
go along. For now, we can notice that given this fact, then one of the things
that’s possible in this slot is, even if the entrant remains silent and somebody
who is pre-present talks, they might not talk to him. They might address one,
or all, of the pre-present persons.

And of course the entrant can, perhaps feeling that he might not get an
invitation if he was just to leave things that way, request one. There are a
variety of ways of doing that. Or he might talk in a way which does not
provide at all for him getting an invitation, and just begin a discussion about
anything. He might simply assume that he has perfect rights to participate,
and not give them a chance to make it an issue.

Now, questions of what might be seen to govern the determination to give
or not give an invitation, are extremely complex. I don’t have any very
comprehensive ways of providing answers. I’ll try to provide a series of
considerations which I take it are required by the data.

One thing I’ve already suggested concerns the matter of getting the entrant
to make a desired identification of the pre-present persons. Via an invitation
they proffer an identification to him of themselves and offer him one also. And
we might say that in exchange for his accepting the identification they proffer
for themselves and for him, he gets asked in - where his acceptance of the
invitation may be seen to accept the identification.

On the issue of the acceptance of a proffered identification involving also an
identification of oneself, we can get at that via a rather general rule about
identifications which goes something like this: If a party makes an identifi-
cation of himself via some category of some categorization device, then the
one to whom it is made, properly returns with an identification from the same
device (which can be the same category). If somebody comes up to somebody
and says, “How do you do, my name is so-and-so, I’m a doctor,’’ the other
returns, “My name is so-and-so, I’m an engineer.” “I come from New
York,” “I come from Chicago,” etc. In this case, again, they’re saying “We’re
teenagers,” i.e., they provide the relevance of the device of which ‘teenagers’
is a category, and he would find that he is, also.

The relevance of that would be, were it the case that there are clearly
alternative identifications, and these alternatives matter. I suggested that
alternative identifications can matter, with the case of the navy pilot as one
instance. In that case, the device of which (military man/military man) was
a category, stood in alternation to a device which contained categories such as
(criminal/law enforcers). Here, the device of which ‘teenagers’ is a category
stands in alternation to the device which contains the categories (therapist/
patients). And I’m going to say that that device is an ‘ omni-relev ant device' .

An ‘omni-relevant device’ is one that is relevant to a setting via the fact that
there are some activities that are known to get done in that setting, that have
no special slot in it, i.e., do not follow any given last occurrence, but when

314

Part 111

they are appropriate, they have priority. Where, further, it is the business of,
say, some single person located via the ‘omni-relevant device,’ to do that, and
the business of others located via that device, to let it get done.

In this case, the new entrant comes in, somebody goes about greeting him
and introducing him:

Ken : Sure. Well let him. You know-

Roger: On Hollywood Boulevard the other night they were giving

tickets for dirty windshields, ((door opens))

Jim : hh

Ther : Hi, Jim / / c’mon in.

Jim : H’warya

Ther : Jim, this is uh Al,

etcetera. Things may be going along, the device isn’t being used; at some
point something happens which makes it appropriate, and it’s used. And
when it’s used, it’s the controlling device, i.e., there is no way of excluding its
operation when it is relevant. You could perhaps locate such things; for
example, the relevance of (teacher/student) in a seminar where everybody is
sitting around talking in a more or less homogenous group, by virtue of who
it is that announces break time, or who determines that there’s going to be a
break or not, somebody else having suggested it. In the case at hand we have
this insertion of priority items - greetings and introductions - by this guy, the
therapist; and the participating in it by these others, the patients. Later on in
the session we get an interchange of the following sort:

Ken : I still can’t figure out what the hell I spilled on my shirt.

Roger : Alright I give up what’s on your shirt?

Ken : “Kill it!” heheh hhhhehh

Al : That’s not loud enough. KILL IT!

Roger : You’re gonna bust this tape recorder.

Several : ((light laughter))

Ther : Well, what’s new, gentlemen?

Al : That’s a hint we must get outta here before he gets mad at

us.

Roger : We adjourn to lunch now.

That is, the closing of the session is something which this same fellow who
invoked the greeting-introduction sequence, also invokes.

One way that we can see that a device (a set of categories and some ways
of using it) is omni-relevant, turns on there being some insertable sequences;
for example, the greeting-introduction sequences and the closing sequences,
for whose orderly use an orientation to that device is essential. The device is
omni-relevant by virtue of the fact that the insertable sequences, while having
various kinds of priorities, don’t occur at any specified point, i.e., any point
located by virtue of, for example, a last occurrence. In regard to this particular

Lecture 6

315

case, that device consisting of the categories (therapist/patients) is omni-
relevant for this two-hour conversation; where that is warranted by reference
to the fact that the greeting-and-introduction sequence and the closing
sequence are controlled by reference to that device. Now, I’m not saying that
(1) the only device that is omni-relevant here or anywhere is the one with the
categories (therapist/patients) - it may well be that there are other devices in
this conversation that are omni-relevant, or that (2) the only way to see that
something is an omni-relevant device is by reference to these insertable
sequences - there are other ways.

Now let me remark on the technique used to see and warrant that
(therapist/patients) is an omni-relevant device here. Take a piece of data like
this little sequence:

Tber\ Well, what’s new gentlemen?

Al : That’s a hint we must get outta here before he gets mad at us.

Roger : We adjourn to lunch now.

We develop some analysis of it. We probably don’t need as interesting a piece
of data as that, but it is what occurred in this case. By ‘interesting’ I mean that
the therapist might have ‘directly’ said something that anybody would
recognize as the beginning of a close, i.e., the introduction of a closing
sequence. In this case he happens to use something ‘cryptic,’ i.e., something
that had to be analyzed to have that found.

It is by virtue of our seeing that this item gets placed by this fellow, and
the others see what it is by virtue of seeing that he’s doing it qua therapist, and
is doing an intended close, and that they inform this new entrant of its import
for them qua patients, that we come to pose our problem. I want to offer for
consideration a difference in method from other ways of going about building
machinery that analyze something or other.

It is a very conventional way to proceed in the social sciences, to propose
that the machinery you use to analyze some data you have, is acceptable if it
is not intendedly the analysis of real phenomena. That is, you can have
machinery which is a ‘valid hypothetical construct,’ and it can analyze
something for you. So you say, for example, ‘I have a bunch of constructed
phenomena that analyze some real thing. I am not saying that people do that
thing via my constructs; only that the apparatus I have will predict. . .’ etc.
Now that’s not what I am intending to do. I intend that the machinery I use
to explain some phenomenon, to characterize how it gets done, is just as real
as the thing I started out to explain. So, for example, if I’m going to use some
piece of data with respect to the issue of omni-relevance, then that piece of
data has to be analyzable itself, and that analysis would have to be correct, to
work, etc., quite independently of its use for omni-relevance. That is to say,
we could analyze that closing sequence, work out some of its features, before
we had any idea that it involves a piece of machinery that would provide for
the participants that there was a phenomenon of omni-relevance for the device
that it involved.

316

Part III

There are some real importances to that sort of thing, apart from general
issues like ‘are hypothetical constructs good things, necessary things, lousy
things,’ etc., which make this procedure attractive to me, anyway. First of all,
I take it that I want to be dealing with the real world, in all of the work. But
second, what it forces is, if you’re going to build a piece of machinery, then,
as you build it you find things out about its parts as well as that it is a part
of something else. So even if it didn’t turn out that this was a piece of
machinery for omni-relevance, it would be a something; we would have
studied it, we might have it available, and we might figure that Members
have it available for uses in one or another place. And recurrently it happens
that some piece of data is analyzed, and when you’re analyzing something else
you find that the machinery built and tested to analyze one thing is now
important for this other thing. That permits you to tie things together.

It also makes for certain difficulties. One of them being that the rate at
which you solve problems is not at all given by the fact that you had some
problem you wanted to solve. So, for example, I had the sense a year ago that
‘therapist/patient’ is an omni-relevant device here. I didn’t know how to
show that this was so, or even how to find that this was so. There were bits
of data that looked like it, but until, as it happened, I was working on these
things without any intention of dealing with the issue of omni-relevance, not
much was happening. It was when I was working on the priority character of
those insertable sequences, and asked, ‘how is that priority invoked, and what
allows this one or that one to do it?’ that it became apparent that it was
material which was very much related to the phenomenon of omni-relevance.
Now that happens in many cases. It means that you often have to wait for
some data to strike you, or to occur in some sequence for you to be able to
solve some ongoing problem. Or you may have solutions sitting around to
problems that you haven’t been able to pose.

But the core point is that when you introduce a piece of machinery, that
piece of machinery in the first instance introduced where it, itself, analyzes the
things of which it’s built, you aren’t then, building some elaborate construct
which, if it doesn’t analyze the things you ‘want’ it to, you’ve got nothing.
You may have learned an enormous amount, as you’ve fitted various pieces
of machinery together, about what they’re doing, about other data, and it
turns out that they just don’t analyze what you expected them to, but there’s
no loss there. In any event, one intends in the first instance, that each piece of
any given apparatus is saying something about some particular real thing.

We have, now, the notion of an ‘omni-relevant device,’ with ‘therapist/
patients’ being one in this case. We can note that this device is adequate to this
environment, i.e., that each of the persons there before and after the entrant
came in, can be categorized by those categories. That it is omni-relevant may
not mean that they thereby must invoke it if, e.g., they’re going to do an
identification. For one, there may be other omni-relevant devices; for another,
there may be other appropriate things to do. But the fact that it is
omni-relevant may set certain constraints on what other identifications they
could use. A question is, what kind of constraints might it set?

Lecture 6

317

The notion I’m going to propose now, in offering a constraint, I’ll give a
term that’s rather generally usable for certain kinds of identification, and I
think that the way I’m using it captures its natural sense. The question we
may be asking when we ask what other kind of device could they use than the
one that’s omni-relevant, might be asked in a context like this: Suppose
there’s some identification that could be applied to a set of persons, some of
whom don’t like that identification for various reasons, but some features of
that identification are important for various reasons. What they might want,
then, is something we could conventionally call a ‘cover.’ Where a ‘cover’ is
an identification which is more palatable, which can hide the problematic one,
and which nonetheless allows whatever it is that the problematic one can do,
to get done. And there are easily thinkable ways in which, for the kids, the
identification ‘patient’ is not the most delightful one. They may be ashamed
of it, etc. And it may be that they want to have, for purposes of the discussion
at least, a cover for that.

One thing that may be necessary for a cover is that however it is that the
omni-relevant device partitions persons, e.g., (therapist/patients), that that is
held constant for the cover device. The notion, then, is partitioning constancy .’
Consider the eligibility of the device that consists of ‘stage of life’ (adults,
teenagers, etc.). It could be eligible if those partitioned into one category in
therapist-patients were together in adults-teenagers, and those partitioned
into the other category also remained together. Then they might introduce
themselves as a bunch of people consisting of adult and teenagers, where
various things that the (‘therapist’/'adult’) does and that the (‘patients’/
‘teenagers’) do among themselves could be perfectly well handled; they would
all be in the right boxes together for the purposes of the omni-relevant
category (therapist/patients).

Now, if there are devices which, for a given population, have ‘partitioning
constancy’ with an omni-relevant device, then it may well be that those are
usable instead of the other, for purposes of identification. And if they are, and
the other is still in operation, we would talk of them as a ‘cover identification.’

One of the problems in constructing a ‘cover’ is that it permit you to have
a set of activities look like they’re perfectly ordinary. For example, suppose a
mental patient is going into town with a nurse. What they want is to come
off, not as a mental patient and a nurse, but as something else which is not
noticeable, perfectly ordinary, acceptable, anybody could be in it, like
‘male-female.’ Then that would provide a set of constraints on how they
might behave. It would mean, for example, that she doesn’t lead him around;
that’s not the position of ‘females.’

And it may well be that you can only pick certain devices to be able to do
what you want in more or less public settings while preserving certain
relevancies - whatever they happen to be - for the things that involve you in
doing what you’re doing - like, in the case at hand, engaging in therapy. In
that regard, you might consider the problems of spies. Spies have to build
‘covers’ all the time. They have to build covers which provide for them seeing
certain people and not others, doing various jobs, buying various things, etc.

318

Part III

They want to pick something that permits them to do what they have to do
while not making that thing noticeable.

That the kids, or at least the one who says “We were in an automobile
discussion,” is building a cover, could be gotten via a consideration in general
of their notion of what the entrant might expect them to be doing in a group
therapy session. They’re proposing it is simply a matter of conversation that
perhaps any teenagers could have. Nothing special at all is going on. Whereas
the entrant might imagine anything is going on; it’s a bunch of kids who may
well be more or less ‘crazy.’ He might not have any good idea what in the
world these people are doing, and in fact, it looks like they figure that that’s
so. About ten minutes after the session has begun, the following interchange
takes place:

Al : What time is it?

Ther: Ten thirty.

Al : What happened to Looey. Is she not coming this week?

Ther : She- no she won’t be coming in.

Al : Where’s Jim Reed?

Ther : Jim Reed will be late.

Roger : Who’s Jim Reed?

Al: New guy is coming.

They then proceed to do a set of ‘joking’ greetings and invitations:

Roger : Aaaghooh! ehah hah hah heh heh

Ken : Smoke him out!

Al: heh Hi, Hi,

Roger : Put cher shoes on! hehhehh

Al : Hello, Jim, take a whiff. This is my pipe,

Ken : hehhhehhhehh

Roger: ‘ey we got a homey atmosphere in here hehhehh Lay down, heh

heh

Al : ehhah!

Ken : hehhehh ( )?

Al : Yeup. Walk in, take your clothes off, and be ready.

Now, that’s rather different than “We were in an automobile discussion,”
and they might take it that Jim would expect something like the above series.

Another consideration is, it’s perfectly clear that their ‘mental states’ are the
thing that is providing their relevant presence. Not any teenager is a member.
So one thing is, perhaps, the issue of alternative devices as involving some
basis for offering an invitation which would, by virtue of this identification tie,
do the job of inviting him to use, and see as usable, this ‘teenager/adult’
device.

We could also consider the following problem. After all, this fellow was
processed by the clinic in the first instance; that is to say, he or his parents said

Lecture 6

319

that he ought to go there, they gave him whatever tests and decided that he
ought to come in, and the management assigned him to some particular
group. And now the question is, do the members of that group, i.e., the
patients, have any say whatsoever as to his acceptance, or is it the case that
having been selected by the management, that’s it. They have no business in
the first instance, even suggesting that they have to make an invitation, or
they can, or they ought, or in any event that it’s relevant, but if it’s relevant
that they better come out with one right off, and they have no further say-so
over his acceptance.

The issue of an invitation, then, can be concerned with whatever challenge
they’re going to make to the authority of the management, either via the fact
that they see that they have rights to make an invitation, or, more importantly
as it will turn out, that they don’t give one. And they explicitly take it that
his membership has to be developed. Management can bring him in, i.e., say
‘Come in,’ but can’t make him a member of the group, i.e., say ‘Join in.’

Lecture 7

Cover topics ; Collaborative sentences;
Tying rules; Relational-pair

identifications

I want to begin with a further remark on the matter of an identification being
offered which might be a ‘cover.’ First of all one might notice that when an
explicit invitation is offered to this fellow, it isn’t “Would you like to join our
group?” or . .this group?” or “. . .this therapy group?” but it is “Wanna
join our army?” Again, it’s an alternative category to the one that is relevant
to - indeed, controlling of - his presence, at least from their point of view, to
some considerable extent.

In line with the issue of a possible ‘cover’ identification, one might consider
this topic that gets proposed, ‘an automobile discussion.’ It is, for teenage
boys, a very special kind of topic; one which is perhaps as ideal a one as could
be had if one wanted not simply cover identification but a 'cover topic.’ and
that’s because of the range of matters which are expectable and in fact
discussed at a place such as this, i.e., therapy topics, which can be discussed
under the guise of ‘an automobile discussion:’ Sex, guilt, independence,
autonomy, authority, parental relations, the state of society, death, you
name it. All can perfectly well be handled as sub-topics of ‘an automobile
discussion.’ And even in our very short segment, more than a few of them
are.

The automobile discussion, then, can provide perfectly ordinary, i.e.,
things-that-teenagers-do, bases for dealing with those matters rather than, for
example, their being introduced by virtue of the fact that they are the proper
things to be talking about in a therapy session, or they are the things that
independently disturb these kids. And it might not be simply incidental that,
if the format of ‘an automobile discussion’ is used, then there’s a very nice
shift that takes place in the relationship of the two sets of persons present (the
adult therapist and the teenage-boy patients), and that is, that on this topic
they’re the experts and not he. (Though we might say that on each of the
sub-topics he ought to be the expert.) And one might explore that also, as
further related to this issue of offering the cover, in terms of which anything
that gets said could be understood as properly raisable, without noticing that
such matters would perhaps have to be dealt with anyway for - as they put
it - “the cure” to go on.

When I started considering the line, “We were in an automobile
discussion,” I said that it is possibly to be seen as an invitation, were it to occur

320

Lecture 7 321

alone. As it happens, it does not occur alone, and I want now to consider the
actual features of its occurrence.

Ken : We were in an automobile discussion

Roger : discussing the psychological motives for

( ): hhhhhhh

Al : drag racing on the street.

Really a most remarkable phenomenon. Three people collaborate in the
production of a single sentence. Now, collaborative productions of a single
sentence are to some extent not terribly rare. But those that are not rare are
relevant in the way they indicate how special the phenomenon is, otherwise.
For example, you’re sitting at the table, somebody makes a request: “Pass the
salad.’’ Somebody else, say, perhaps, the person who got the request, appends
to it some word or phrase that they propose properly completes it: “-please.”
Another way ‘collaborative’ utterances get done is, say, one person produces
an almost complete sentence and finds himself searching for a last word or a
last phrase which he can’t find. Another offers it. So, in one of the therapy
sessions we get:

Ken : He looked like he was coming in here for uh. . .

Al : -kicks.

Louise : -guidance, heh

Note here that the inserted word is treated as a ‘candidate,’ and the other
person then accepts or rejects the candidate term, and goes on to complete the
utterance. (In the case at hand, Ken rejects the terms offered, and continues
to search for the phrase, for approximately another full minute, and there are
further offerings of candidate phrases and terms.)

And there is, in the segment we have, an example of another, very
characteristic two-party production of a sentence. It turns on the idiomatic
phrase, ‘circumstances beyond one’s control:’

Roger : Kids don’t drive long. . .by the time they’re eighteen they’re bad

walk(hh)ing hehh through circumsta(hh)nces hehhehh

Ther : beyond their control.

Roger : ehhehh Yes hehhehh

But for sentences of the complexity and un-idiomaticness of form of ‘ ‘We
were in an automobile discussion. . .etc.,” I take it that it’s really a striking
phenomenon.

Now, one obvious thing that such a production might be seen to do - and
it’s perhaps as neat a conceivable means as you could have for doing it - is
showing the new guy that this is a group. They do together what is among the
most prototypical things that a single person would do. And they do it in close
coordination; there are no pauses that are much noticeable in that sentence.

322

Part 111

I tried to find some literary uses, but it’s so powerful it’s banal. It would
just be corny. I heard a musical version of Edgar Lee Master’s Spoon River
Anthology , where the device was used, and it was used to do just exactly the
same task. He was trying to show a pair of happy lovers, and the happy lovers
talked this way, e.g., she would begin a sentence and randomly stop in the
middle, and he would continue it. And that way they provided a demon-
stration of Freud’s remark that only in love do we feel we are one. And I’m
told that in the comic strip Donald Duck, the author puts together three of
the character’s sentences in that fashion.

We get, then, a kind of extraordinary tie between syntactic possibilities and
phenomena like social organization, i.e., that way of not stating, but
presenting the fact of organization. There probably isn’t any better way of
presenting the fact that ‘we are a group’ than by building a new sentence
together; one that’s coherent, grammatical, etc., and unplanned. Its produc-
tion involves, very markedly, a very very close and quick attention to the talk
of each other. And in considering this usage we’re going to begin to develop
some extremely central matters. From it, we get the kernel of what makes
something that could be called ‘a conversation,’ in the sense of how it is put
together. And one thing that involves is a long list of things I’m going to call
‘tying rules.' They are a means by which one piece of conversation is tied to
another.

If conversation simply consisted of A-B-A-B in alternation, then one
might, for example, be perfectly well able to disorder all the parts, as long as
the alternation is preserved, and still have a recognizable conversation, or even
the same conversation. What these ‘tying rules’ do is radically restrict that
possibility, and provide for very local control over the relationship between
utterances.

Now, what happens with “We were in an automobile discussion” is an
example of a ‘tying rule,’ and with that example we will have hundreds of
examples constructable:

Ken : We were in an automobile discussion

Roger : discussing. . .

One of the things we see that’s interesting is that “We were in an automobile
discussion” is a perfectly good sentence by itself. When we get “discussing the
psychological motives for,” what we have is something produced which, by
itself is not a sentence. Its user makes his piece of talk a dependent clause and
makes the prior piece of talk an independent clause and not a sentence. So he
hooks his statement onto the previous one.

This happens to be but one case of one large class of these tying rules; and
these are what I’m going to call ‘second speaker’s rules.’ What I mean by
‘second speaker’s rules’ is that a speaker makes himself a ‘second speaker’ by
something he does, and makes, then, the one who spoke before, a ‘first
speaker.’ He does that by the way he ties his remark to the one that’s gone
before. So, for the analyst, cases of the second speaker rule are not found via

Lecture 7

323

the fact that one sees some utterance in the first instance as a ‘first utterance,’
made by a ‘first speaker.’ You start with something recognizable as ‘the
second member of a pair,’ and then turn backwards to something that is not
recognizable as a first member of a pair, unless one has the second to see that
there ought to be one. And those are in contrast to first speaker tying rules,
e.g., I ask a question. I make myself a first speaker and the other, a second,
where the first item is recognizable as ‘a first item in some pair.’

What we can see is that the rules for building dependent clauses are second
speaker’ rules. In the case of “We were in an automobile discussion,’’
“discussing. . we could say we have a rule which says ‘if you want to make
yourself a second speaker by building your statement as a dependent clause to
a prior sentence, one that could be constructed as an “independent clause,”
then, if that one ends in a noun, participialize the noun and talk in some
syntactically consistent way.’ For example, “We were having dinner.” “dining
on roast beef. ’ ’ Of course you can use whatever else the grammar provides as
ways of building dependent clauses.

Now, if we consider this second utterance which ties itself to the first, i.e.,
“discussing the psychological motives for,” one of the things we might want
to know about it is, what would be relevant to tying the utterance to the first
and not simply beginning a second sentence, e.g., “We were in an automobile
discussion” followed by a second sentence, “We were discussing the psycho-
logical motives for drag racing on the street.” What I have to say is not
intended as exclusively or exhaustively an answer to that, and perhaps it isn’t
even plausible — much less convincing - even as a partial consideration. But
there are some good grounds for suggesting that such things as I’m going to
propose are sometimes relevant somewhere. They concern the following sorts
of things.

I’ve mentioned this matter of paired utterances , like question-answer. They
are of course enormously relevant in two-party conversations. They gain a
larger relevance in what I’ve been talking about as ‘multi-party’ conversa-
tions. They gain that larger relevance because in multi-party conversation
there isn’t a formulaic way of locating who should talk next, i.e., there is no
simple expansion of A-B-A-B to, e.g., A-B-C-D-A-B-C-D. What one gets,
then, is a heavier reliance on these paired utterances which assign the second
person’s rights and obligations to talk; for example, if they’re addressed a
question, then they have an obligation or a right to talk next.

That seems to provide a differentiation in the phenomenon of interruptions
in two- and multi-party conversations. In two-party conversation, if A is
talking along and B starts to talk in the course of it somewhere, then we might
say that B is interrupting A. In multi-party conversation, at least where we’re
dealing with some utterance pair, that doesn’t seem to be the case. What
seems to be the case is something like the following. Suppose A addresses a
question to B, and now C starts talking at the end of that sentence. Then, if
B (the proper answerer) starts talking also, regularly it turns out that while it
would read off the tape like the proper answerer is interrupting C, it isn’t
treated that way. C seems to treat himself as having interrupted B (the one

324

Part III

who hadn’t yet spoken) and stops. That is: A to B; C starts; B starts; C stops.
We get, then, a notion of ‘interruption’ which is partially independent of who
happens to be talking. That is, again, C is talking and B starts to talk. We
might figure that B is interrupting. But C is interrupting from the moment
he begins.

That means in part, say, that if “We were in an automobile discussion”
were an ‘adequate complete utterance,’ and it were an invitation to this fellow
Jim, then it’s the business of these other guys, accepting the fact that the
invitation has been made, to find a way to do what they need to do before Jim
properly speaks - perhaps to accept the invitation. What they do is provide
that an adequate complete utterance has not occurred, so that the slot for Jim
to speak doesn’t open here. They then proceed to modify the invitation, which
can then be heard as: “Hold on for a while, see what we’re talking about, and
how we do it.” That is, for Jim, “automobile discussion” might be a very nice
thing to hear. “Oh, they talk about whatever it is kids talk about anywhere,
and they’re not going to be talking about things that are really godawful
things to talk about.” The modification provides that when the slot does open
up, he has been told that in fact they talk about things that he might not be
able to participate in right off.

Now that sounds like an extravagant orientation to these slots being
opened or not, and were I just to have that, I wouldn’t believe it, either.
Indeed, I wouldn’t have come up with it had I not had other, and very strong,
indications that there is such an orientation to slots. Here are some of them.

First there has to be a differentiation between those kinds of items that have
strong but short control over sequencing, and those that may have strong or
weak, but long control. A thing like question-answer is a case of long control.
That is, if you ask a question, then it may be open to be answered any time
over whatever future as long as it hasn’t yet been answered. And it isn’t
required that it be answered right off, so that you can get other things fitted
in between the question and an answer. A simple one is:

A: Did you put out the garbage?

B: Did I put out the garbage?

A: Yes.

B: Yes.

Some things don’t have that character; for example, greetings. If I say
“Hello,” thenifyou don’t answer with “Hello” or something that’s a member
of the class of ‘greetings,’ but introduce something else, then you wouldn’t at
some indefinite point into the conversation offer “Hello” as the return.

Here is one thing that has a short control, but seemingly a rather powerful
one. I was studying phone conversations between members of a psychiatric
clinic and callers who tended to be suicidal. The conversations began in a very
regular way, on the part of the psychiatric clinic. They would say things like,
“This is Mr Smith, may I help you?” and things like that. There’s a proper
return to that, which was regularly offered, and by non-suicidal persons it was

Lecture 7 325

almost universally offered, which is, to “This is Mr Smith may I help you?,”
“Yes. This is Mr so-and-so.”

Now there are a lot of people who don’t like to give their names. Especially
people who are in one way or another troubled about what they’re getting
into, etc. You could simply propose that all they have to do is just not give
their names; to “This is Mr Smith may I help you?,” just say “Yes” for
example. But they didn’t do that. What they did was to use some more or less
elaborate means which served to never have the place where their name
belongs occur, such that they never gave their names but their names were
never absent, i.e., that slot never occurred.

They used several means which are very generally usable to do a similar
task. One of them was, for example:

A: This is Mr Smith, may I help you?

B: I can’t hear you.

A: I said my name is Smith.

B: Oh Smith.

A: May I help you?

That is, if you ask a question that indicates that you can’t hear what was said,
upon the return to that question, which clarifies or repeats what was said, you
have the right to repeat that thing to indicate that now you’ve heard it. And
in doing so you fill the place where your name might otherwise occur. That
is, this thing having been broken apart, if the other then comes back with the
second part of it, “May I help you?” you can answer that, “Yes,” and your
name is not ‘absent.’

Sometimes that got extremely elaborate, so that it would go on for a
relatively long time. And while there might be some attempt by the staff to
avoid allowing the slot to be closed, the callers normally won out and their
name did not occur and was not absent. That they were doing this, seemed
quite clear. In any event, it got a kind of verification by virtue of the fact that
for people located in this way (and they could be located immediately, right
off) if you asked them their name later on, they wouldn’t give it.

It’s the fact that they go through all this trouble - if you want to call it
that - i.e., they don’t simply ignore that sequential structure - that suggests
to me that some such consideration might be involved here, where “We were
in an automobile discussion,” with its character as an invitation, is not
allowed to close and the slot for the invited person to speak is not allowed to
open until after that perhaps drastic modification has been provided.

I want now to take up a matter related to our considerations of an
identification, raised by a question somebody asked me. The question was,
Am I supposing that determinations of activities are prior to determinations
of identification? And it may seem that that’s what I’ve been saying or
implying or assuming.

Now first of all, I’m not assuming it. I think it’s so for very many cases.
If you want to play baseball then the activity provides the relevance of the

326

Part 111

collection consisting of (first baseman, second baseman, etc.). And there are a
great many activities that provide for the relevance of some device or some
collection. It’s very important that that be so. And I want to deal with the fact
that it’s so.

If what we’re dealing with is identification via membership categorization
device categories, then there is a large problem for persons identifying each
other or somebody else, which is: There are always alternatives. If you’re
going to use categories of a single device, which one do you use? The question
is, is there some more or less general solution to that kind of problem? That
solution has to be one which is usable by persons who don’t know each other,
if, for example, they’re going to identify themselves to each other, say, at the
beginning of a first conversation. They want to convergently select the same
device as relevant, so that one doesn’t start off by naming his religion when
the other is interested in his occupation. And if one can pull out certain ways
of getting devices via activities, that helps matters out. If you know the
activity, you have the way of categorizing a person. If you’re going to have a
party, use ‘males and females,’ etc., etc.

We can try to consider what sorts of general ways we might construct for
arriving at devices for categorizing. Suppose we have an nth person to
identify, and that nth person is not a first. Then there’s a routine solution: If
the others have all been categorized by some given device, use the consistency
rule to get you a category from the same device. But if you’re dealing with a
single person, then there doesn’t seem to be any unique solution, where a
unique solution might be the basis for a convergent solution. That is, if there’s
a unique device which is appropriate in one way or another, then that might
be the way that persons get together independently on identifications.
Obviously for some activities there’s a unique solution; a device bound to that
activity.

Now, here’s a piece of data:

A: ((a little girl)) Who’s that?

B: ((her mother)) That’s Rita. Remember when you went to the party

last week and met Una? Well that’s Una’s mother.

That’s not only a very characteristic way that persons are identified by others,
but it’s a very characteristic way that people go about identifying themselves
to each other: “Do you know X?’’ “Yes.” “Well I’m his neighbor.”

In considering this piece of data and its character, one can perhaps come up
with a rather elegant solution to the problem of doing identifying, where
there’s no other activity being done than identifying. So you can’t find an
activity which solves the problem of how to choose a device for you. You’re
simply doing identifying, and if you’ve got one person, you’ve got a problem.
A great many devices are available. Now consider the following possibility.
We know there’s a solution for an nth person. Is it the case that we can get
a solution for two persons? Yes. There’s a device that can be used to categorize
a population of any size and it consists of a set of ‘relational pairs;’ things like

Lecture 7

327

(father-son, brother-brother, sister-sister, friend-friend, neighbor-neighbor,
stranger-stranger. . .etc.). Any two persons can be categorized by it. If that’s
so, then you have a way of solving the single-person problem, which is: Take
the person you have to categorize; treat them as the second person of a pair
for which the first is known; find a first. If you can find a first - and in
principle you can find a first - you’ve got a solution. If you can’t, then you’re
no worse off than you were to start with, i.e., you’re back with the one-person
problem which is not capable of giving a solution in any event.

Note that the device that consists of ‘relational pairs’ is uniquely applicable
to two persons. That is, for any two persons there is a device consisting of
‘relational pairs’ which will provide for them to be categorized as a pair - and
only one such device that could hold for any two persons, i.e., it’s a unique
device. And the two-person problem being solvable uniquely, may well be at
least part of the basis for the tremendous regularity with which it gets used.
Where, routinely it’s the case - and it’s the case in this piece of data - that it
isn’t a matter of some first person being chosen with whom you have some
special intimacy. It’s not that you choose only those persons as first of a pair,
who really matter. In this case, the kid barely knows Una, and nonetheless the
mother apparently takes it - and we regularly take it - that when we’ve got
such a pair, the identification is complete.

Again, we can come to see ‘uniqueness’ as relevant in that, in the first
instance, if an activity does have a unique device tied to it, then the device is
regularly used. And also, recognizing that a name is not treated as an
adequate identifier, but when any category, perhaps, is offered, the identifi-
cation is treated as adequate. If the world works that way, that’s extraordi-
narily lovely.

Orientation;

Lecture 08

Being ‘phoney;’ Hinting

I want now to make a few observations about the possible orientative, as well
as the possible identificatory characteristics of this line, “We were in an
automobile discussion.”

On the matter of orientations: In the first instance, it’s perfectly plain that
one way you can orient people to, say, what was going on, or what it is you
do, etc., is by the name of some activity. Now what becomes interesting about
an orientation, for one, is the question of its sufficiency. If somebody is doing
what seems to be an orientation, then they can regularly enough use either a
name that does orient the other person, or one that doesn’t, both of them
being perfectly correct. Suppose you turn to somebody sitting next to you at
a dinner party and say, “What do you do?” and they say, “I do assaying with
an ultra-centrifuge.” And you say, “What’s that?” Then they might say,
“I’m a lab technician.” If, however, it was a group of lab technicians, then it
would be quite sufficient to have said, what, specifically, you do, and more
relevant than ‘lab technician.’

One feature of orientations, then, is the way their sufficiency turns on an
initial identification made of the recipient by the one who is doing the
orientation - where that identification can be categorial. On the basis of that
identification, one would build what one takes to be a ‘sufficient orientation.’

The core character of that initial identification of the recipient can easily
enough be seen in considering how you make, e.g., ‘put down’ orientations.
Suppose four of you are sitting around the table and a fifth comes in and you
say, “We’re playing bridge. That’s a card game for four people. . .” and you
go on and give a detailed description, where you know damn well that
“We’re playing bridge” is perfectly sufficient. Alternatively, you can construct
orientations which you know the other will not understand, and you can do
that for a variety of reasons. Say you want to check out whether they’ll go
along with anything.

The question “What’s that?,” say, after “I do assaying with an ultra-
centrifuge,” is a remark that ‘the identification you used on me to give the
sufficiency for the orientation you made, was wrong. ’ And it can be wrong in
a variety of ways. For example, it could be wrong by virtue of the fact that
you properly used some hint that I seem to give off, e.g., that I’m sitting next

This lecture consists of Fall 1966, lecture 8, with a segment of Fall 1966, lecture
3. Again, it deals with matters not discussed in the spring 19 66 lectures, and it has
been placed where it ‘belongs.’ But in fact, these matters make their first appearance
in the Fall 1966 lectures.

328

Lecture 08 329

to you at a dinner of lab technicians, but it turns out I’m only a friend of some
lab technician and not one myself.

Again, then: One rather important aspect of doing an orientation is the
issue of sufficiency. And then there is the fact that designing the orientative
remark with its sufficiency in mind, involves having made some identification
of the intended recipient.

Now, for the line “We were in an automobile discussion,’’ it is perfectly
possible for such orientative remarks, placed as this one is, to have no
identificatory import in regards to those who make them. That is, it need not
be an identification made via an orientation. They might be things that, for
one, ‘merely happen to be correct,’ i.e., in fact, before this guy came in they
were engaged in an automobile discussion. And if the phrase ‘an automobile
discussion’ is sufficient, their saying it simply clues him in to what it was they
were doing. And especially when one considers orientatives other than
remarks such as this one, there isn’t any necessity that they be combined with
an identification.

But the issue I’m going to address is, what other matters are involved when
the orientatives are possibly involved with an identification. Where, putting
it roughly, let’s say, for example, that one names as what one was doing,
something that may serve as an orientation of itself, but the something that
one was doing provides a ‘hint,’ say, as to the sort one is. And this may have
some important feedback on what kinds of things you can use to make an
orientation where you don't want to make an identification.

Now an added issue comes up when an orientation is being used to do an
identification. That issue is relevant in seeing what you could use to make, say,
an orientative identification, or whether you can use something to make an
orientation alone, without it implying that you are also making an identifi-
cation. The issue is that which we can call ‘misleadingness.’ And closely
related to it is the notion of being ‘a phoney.’

One of the first things involved in the phenomenon of doing ‘being a
phoney’ is that you can perfectly well be accused of being phoney without
having made any explicit claims to being a member of a category that you’re
seen to be not a member of, though having, indeed, intended to suggest that
you were. That one can be seen to be a phoney without having made an
explicit identificatory claim, is, of course, independent of, and a complemen-
tary consideration on, the ways that one can hint at an identity, or be seen to
be hinting at an identity. What one needs to examine, then, are a set of things
one hints with, or can be seen to be hinting with.

So, for example, the magazines on somebody’s coffee table are routinely
seen to be intended to suggest that they are intellectuals, or whatever else.
And you can walk out of a house and say that somebody’s a phoney by virtue
of some lack of fit between what you figured you could infer from various
things in their house, and what you’ve found out about them other than that.
It’s not merely that you say that there is various inconsistent information
being given, or that “At first I thought such-and-such but then it turned out
no,” but one can assign ‘being phoney’ to their use of things recognizable as

330

Part III

‘hints.’ Which is to say, for example, that if they put some magazine on their
coffee table, if they have some painting on the wall, if they use some lingo in
their conversation, then they’re responsible for it being seen as ‘identificatory’
and, if in some way it’s now found to be incorrect, then the assertion ‘phoney’
is properly makeable.

Now there are some rather important methodological issues involved in the
existence of and usability of a notion like ‘phoney.’ Let me just allude to them.
What’s involved are problems dealing with such matters as: Consistency
between a set of facts that you find; attribution of error; when it is that some
problem that gets posed can be considered to be resolved, and a few other
things. And what’s methodologically involved is this sort of thing: Consider
the case of the magazine on the coffee table. A person who comes into the
house can look at it, treat it as an intended identification, and that
identification is now a claim. If, in due course, there are activities which seem
relevant to that claim that turn out to be inconsistent with the initial
determination, then there are a variety of things that, at least in principle,
could occur. With regard to the notion of ‘phoney,’ what’s involved is that
one would not take it that one’s first impression was wrong and got corrected,
though of course that’s a quite legitimate procedure in other circumstances:
“At first I thought he was very cold, but then after getting to know him, I
found out that he’s really a very warm fellow,” for example. Rather, one’s
first impression here - and, very important to the whole thing, without any
claim having been explicitly made - is taken as something that they invite or
require you to make, by what they’ve done.

And involved in that, then, is that you are quite legitimately permitted to
figure that persons know that some object is a ‘hintable with’ object. The issue
of not knowing that something is to be hinted with, although ancillary to our
consideration here, has an independently prominent, interesting place, some-
what largely in psychiatry. There, a claimed naivete about hints, which the
psychiatrist, in company with ‘the rest of society’ can hardly believe, or may
believe but figure it’s amazing, constitute a strong set of symptoms for things
like ‘latency complaints.’

Latent homosexuality involves doing those things which anybody, homo-
sexual or not, knows to be inviting homosexual contacts, let’s say. Where,
when it happens, you’re surprised. And there are lots of things like that. Girls
who get regularly raped. Now I’m not offering an analysis of the psychiatric
situation, I’m only providing the structural place that the phenomenon has,
i.e., the character of certain things as warranting being seen as ‘hinting.’ And
I’m not saying whether they ‘really know,’ they ‘don’t really know,’ etc., or
why they pick up these ‘hintable withs.’

When the thing on the table is seen as a thing with which Members hint
and thereby make a claim to membership, the issue is no longer ‘is he one of
them or not,’ i.e., ‘this is the first piece of evidence, let’s consider the rest,’ but
this is something that he can be held to, such that if it turns out that he isn’t
one of them, you’re not ‘wrong,’ the matter isn’t ‘corrected;’ rather, it now
gives you a different identification whose charactaristics are that it’s based on

Lecture 08

331

the original one having been ‘misleading,’ i.e., ‘a phoney.’

One of the beautiful things about the notion of ‘phoney’ is that you can be
the same category, i.e., a phoney, for a whole range of different possible
intended identifications. You could be a phoney hotrodder, a phoney
intellectual, a phoney beatnick, etc., where that conclusion is in the first
instance a systematic way of processing data, i.e., in which the ‘error’ is now
assigned to the one who gave what is to be treated as a ‘hint.’ And
furthermore, it is taken that they gave the hint intending to have you think
that they were a member. And all this, of course, without explicitly
determining that they know it’s a hint, etc.

What we’re seeing, among other things, is that the time at which some
piece of information is garnered is criterial to the kind of use it has. That is
to say, at that point when an identification might be present, if something is
warrantably seen as such, then a claimed identification is findable. It can then
be dealt with, with respect to whether it stands up. And if it doesn’t, it isn’t
merely that ‘I was wrong about what I thought he was, and now I don’t know
what he is,’ for example, but ‘Now I know what he is; a phoney.’

Now what that has as its import for things like “We were in an automobile
discussion’’ is, unless you’re prepared to sustain what would be seen as a claim
of some identification, then even though it’s perfectly true that, for example,
you were in an automobile discussion, or you did read that issue of the
Saturday Review, or Fortune, or whatever it may be, then you’d better well
put it away, or not refer to it.

And again, the fact of ‘correctness’ gains a kind of irrelevance. If they
‘merely happened’ to be engaged in an automobile discussion, but, e.g., that
isn’t the thing they regularly do, or whatever it is that would be inferred from
hearing that they were engaged in an automobile discussion as something that
would suggest an identification, were not correct, then they’d better well not
say that they were engaged in an automobile discussion. As a possible
identification, then, an automobile discussion is ‘just one thing that such
people do.’ Then, unless, e.g., the other things that such people as are
identified by it do, are done by you, you will have been seen as being
misleading, and thereby, perhaps, also a phoney.

There are other things like: It’s not merely the case that you were in an
automobile discussion, but that you characteristically do it. And that issue of
‘characteristic’ with regard to some ‘hintable’ is itself independently treatable.
Suppose you’re sitting at home one night, you flick on the TV set, and now
you’re watching some serial. There is a knock at the door, and the first thing
you may do is turn off the TV. You may do that by virtue of the fact that
somebody who comes in is going to figure that that’s the sort of thing you do,
and not that you happened to have turned it on just now, though you may
well say, “I just turned it on;” it being a thing which can be characteristic of
some people. It’s often extremely hard to establish some argument that the
thing you were doing is something you did only once. And precisely those
things for which it’s extremely hard to do it, are those sorts of things which
have this identificatory import. That there are a class of such things - and of

332

Part III

course they’re not just statements, they’re a much larger set of things - which
are properly seen as ‘claiming identification’ not only makes them available
for such use, but constrains the various kinds of information, remarks, things
you can do otherwise.

I bring up these issues of ‘sufficiency’ and ‘hintables’ in order to get focus
on what else the various persons who could be seen as being ‘talked on behalf
of (i.e. , Roger and Al), might see implicated in that statement “We were in
an automobile discussion” which they might not want to have happen. And
also to get us in a position to deal with what it is they do with that statement.
They modify it, that’s for sure, and some issues are, then, what the character
of this modification is, and what is involved in the fact that they do modify
it - what ‘are’ they, for one? - and how the modification of the statement
might operate on the possible identification it might have affected. They
don’t, in this case, have the chance to pick up the magazine when they notice
who’s coming in the door, and stick it under the table. The talk is out, and
it has to be dealt with.

Lecture 8

‘We’; Category-bound activities

I want this time to focus on the word ‘we’ in ‘We were in an automobile
discussion.” In The First Five Minutes , you will recall the authors’ analysis of
the utterance “What brings you here?:”

Three of the four words in the sentence; ‘what,’ ‘you,’ ‘here,’ are what
the grammarian calls ‘substitutes’ or ‘shifters’ or ‘pronominal forms,’
words the denotation of which in a specific context depends almost
wholly on that context.

It’s perfectly true that grammarians call them things like ‘shifters,’ etc.
However, it would be rather difficult to support the notion that grammarians
say that the denotation shifts in every or many uses, since that’s not a question
which grammarians address. But it is true that those terms have been
subjected to very considerable discussion by philosophers, and I’ll mention
some of the literature on them.

Nelson Goodman, in The Structure of Appearance, calls them ‘indicator
terms. ’ Hans Reichenbach, in Elements of Symbolic Logic, chapter 7, calls them
‘token reflexors.’ Russell has a discussion in An Inquiry into Meaning and
Truth, chapter 7. He calls them ‘egocentric particulars.’ Quine, in his book
Word and Object, section 21, calls them ‘indicators,’ also. Bar Hillel, in Mind,
63, pages 359-379, calls them ‘indexical expressions.’ Cullenson, in Lan-
guage Monographs , 17, calls them ‘indicators.’ They’re rather fundamental for
logicians because - to use Bar Hillel’s observation - although it may well be
the case that over 95 percent of all natural sentences contain them, they’re
very very troublesome, if not impossible, to handle, if what you want to do
is build characterizations of the truth or falsity of particular sentences.

The core thing about these terms that have been examined, and why it’s of
interest to the logician, is that they have an extraordinary transiency of
reference. So the issue is, if you’re dealing with the truth or falsity of
propositions, then how are you to interpret a proposition which has the term
‘now’ or ‘you’ or ‘it’ or ‘we’ in it? That is the logician’s problem. What we
have to do is to try to construct what a procedure might be for determining
what it is that’s being referred to when somebody says ‘you,’ ‘we,’ etc. These
terms are extremely complicated, and in concerning myself with ‘we,’ while
one point I have to make can be said in a sentence, its consequences wouldn’t
be seen if I said that sentence, so I’m going to go about making the points I

Includes pp. 1-8 of Spring 1966, lecture 9.

333

334

Part III

want to make in what may seem a rather elaborate fashion. The roughest
message you might pick out of what I shall say is that in dealing analytically
with conversation, you must at least be cautious in the use of what you’ve
been taught about grammar.

It is perfectly legitimate, in English, for a single speaker to produce
instances of the following two sorts of sentences: (1) ‘We do X,’ and (2) ‘I
don't do X.’ For example, “We are fickle,” “I am not fickle.” “We cheat,”
“I don’t cheat.”

It might seem that if a single speaker can make both those statements we
can easily derive a contradiction and thereby establish another way (‘another’
in that many have been established) in which English can be seen to be
‘inconsistent.’ They may look, right off, to be contradictory, but it would pay,
even if it looks very routine, to work out how they are, or how they might be
found to be so. For that will then permit us to examine what it is that we do
to find it ‘contradictory,’ and then to see whether what we do is, itself,
independently legitimate.

We would start off by asking, about “We were in an automobile
discussion,” is it the case that ‘we’ is a set containing these guys’ names,
directly? It is perfectly proper to say ‘we’ referring to some set consisting of a
list of members. In the data, Roger at one point says to Al, “We’re doing
better than he is. Proceed, I’m with you.” Where “I” [Roger] and “you” [Al]
add up to “we” [Roger, Al]. This use of ‘we’ is what we could call a ‘set
name,’ and as such, we could have a set with members of the following sort:
One member at least would be ‘I’, and there would be others perhaps on
some occasion, ‘A’ and ‘B,’ where ‘A’ and ‘B’ are now names of names; i.e.,
you could replace the letters with particular names.

We could proceed in two different ways. One of them is: There are
conditions under which if something is true for the set name - or false, for
that matter - then the set name can be replaced by the set members. So we
would get, then: “We do X;” if that is so, it’s so for each of the members,
then we get: [“I do X,” “A does X,” “B does X”]. Now, each of them have
to be correct for the other to be correct, so we do have a contradiction between
(1) “We do X” and (2) “I don’t do X,” since the ‘I’ is contained in the set
‘We do X.” That means that either (1) is false, or (2) is false, or something’s
wrong, i.e., if it’s the case that the language permits both, then in harboring
a contradiction like that, it’s inconsistent.

There is, however, a use of ‘we’ which is quite different, that works in the
following way. If you’ve used any membership categorization device category,
i.e., any category like male-female, Negro-white, old-young, first baseman-
. . .second baseman. . .catcher, fullback. . .halfback. . .etc., doctor. . .law-
yer. . .etc., native-foreigner, any of these, you can on some next occasion
wherein you want to refer to the same object, use a pronoun to do it. If you’ve
referred to a category in its plural form, e.g., Negroes, women, then you
choose from a plural pronoun, most particularly ‘we’ or ‘they,’ and you may
pick ‘we’ or ‘they’ by reference to whether you are, or propose to be, a
member of that category. So you can say, “kids race cars,” “we race cars,”

Lecture 8

335

“kids race cars, we don’t get caught, however.’’

Now the question is, what are the conditions for the correctness of a
statement that would be asserted with a categorial term as the term which is
the subject of the sentence, where any of those can be reasserted with pronouns
in that slot? For example, “women are fickle,” “we are fickle,” “they are
fickle.”

For one, it is the case about such categories (and formulations of knowledge
and facts about them) that the categories or the legitimate plural replacements
may be perfectly properly used without the user having any ability whatsoever
to replace it with a list of those persons about whom it is true. That is, ‘we,’
‘they’ can refer to a category which has as one of its crucial properties that no
intention exists of listing the incumbents, and furthermore they’re not listable,
i.e., an infinite population. Consider perfectly ordinary uses of ‘we’ of the
following sort. The first quote is from the New York Tunes, an article on
Yugoslavia:

Yet Communist officials appear to be optimistic. “We haven’t had good
roads since Roman times, so why should we worry if the highway to
Belgrade is not finished tomorrow”, one said.

The next one is from an article on Israel:

“Don’t forget this is our country,” Ben Josef said with a smile, “We
have been here much longer than since the thirties. Up there,” he
gestured toward the hills, “stood one of Solomon’s cities.”

That is, ‘we’ can be used to refer to all members of the category that have ever
lived and may ever live.

Furthermore, once you’ve gotten the plural pronoun replacing the catego-
rial term, there is no movement, necessarily, to any set-of-members reference,
i.e., reference via a list of particular names. The categorial reference, once
established, holds. It can be said: “Americans do X,” “We do X,” “I do X.”
It can also perfectly well be said, “Americans do X,” “We do X,” “I don’t
do X.” No contradiction at all.

If, however, ‘we’ is being used to refer to a set of persons who are seen as
a list, then, if one has ‘we do X” or “we did X,” and either ‘I’ or ‘you’ who
are on the list haven’t, then one of the two is indeed wrong. And somebody
has a good complaint, and can raise it, and they do: “You just said ‘we do X,’
but you didn’t do it, so how can you say that?”

Given the ambiguity of these terms, i.e., how it is that the use of ‘we’ is
derived, there is the basis for making jokes. For example, somebody reports
some ‘we’ statement which, if it were heard as an assertion of a member of a
list and not a member of a category, would be nonsense: A little boy
announces, “We’re playing the Rams today!” You say, “Are you going in as
fullback or tackle?”

In any event, for any use of ‘we,’ one has to have some way of making a

336

Part III

determination as to, is it being used as a replacement for some category, or is
it being used as the name of a set consisting of a list of persons? Quite different
treatments will occur, depending on which is heard, and the treatments would
be quite anomalous unless one comes to have some way of seeing which use
it is. And of course there can be overlapping use, where the category may well
be one which has a list, and one’s name appears on that list. A member of a
football team can then say in both senses, “We’re playing the Rams today.’’

I want to turn now to one thing that is special about those categories. Such
knowledge as is recognized as correct by virtue of the combination: [a category
(whatever it is) plus some thing they do or attribute they have (category
bound in that sense)}, has a rather important property which I talk of as
‘knowledge protected against induction .’ By that I mean, if you have a
statement, “Women are fickle,” then it is no way to undercut that statement
to introduce as a possible contradiction some statement which consists of a
name of a person, plus not-fickle, where that person is a member of the
category. That is, there’s no contradiction between “Women are fickle” and
“Mary is not fickle,” where, for the latter you can get, if Mary is talking, “I’m
not fickle,” or if somebody is talking to her, “You’re not fickle.”

There are some very important consequences of that knowledge, which is
in the first instance formulated by one of the membership categorization
device categories as correct, being protected against induction. And that it’s
not merely an assertion of mine that it has that character can be easily enough
seen by considering how one understands things like the following. Suppose
some positive piece of information is bound to a category you’re a member of,
e.g., “Jews are rich.” If a member of that category is not, then he tends, not
to see himself as a living demonstration of the error of that statement, but to
take it that there’s something the matter with him. And one of the important
things about ‘knowledge protected against induction’ is that members who
find for some of it that it doesn’t hold for them, have a program laid out,
which is to make it true for them. Furthermore, for some such things they
may come to feel, again, not that the knowledge is wrong, but that they’re
‘phonies;’ they’re not real members. That is, one way persons come to see that
they’re ‘phonies’ is by virtue of some piece of knowledge, holding about a
category they belong to, does not hold about themself.

As another way to see that there is a corpus of knowledge which is
protected against induction in the way I talk about, consider the following.
The first one, the first five, the first ten members of some category that you
meet, you can perfectly well see as ‘exceptions.’ You don’t say, “See? What
we know about Negroes is wrong,” but, “I’ve met many exceptional
Negroes,” for example. And that holds for each of them. The fact that you
see a person as ‘an exception’ involves, then, a way that you use this
knowledge protected against induction.

Note that knowledge protected against induction is not simply important
as ‘maintaining negative stereotypes,’ if you want to think of it that way, but
consider, for example, if something is known to be so about an occupational
category, or, for that matter, a religious category, then the list of things which

Lecture 8

337

are known to be so about them provide an advertisement for persons who, if
they want those things said about them, or want to feel those things as so
about themselves, have to arrange their lives so as to become members of this
or that category.

Let me try in perhaps a little different way to bring home a few points
concerning the character of statements like “Women are fickle,’’ “Jews are
rich,’’ etc. which it appears to me many people don’t properly appreciate, i.e.,
understand. I’m sure that many people, especially the more liberated, think
that those are stereotyped statements which aren’t really true; they are used,
and those who use them are unfortunate objects of contempt or subjects for
better education.

Holding the notion that they’re not true involves hearing them in the first
place as candidates for being true under some formulation such as: What
they’re saying is, ‘women’ means ‘all women;’ that’s subject to a test; for
example, go out and use the device ‘male-female;’ locate all cases of the
second class and determine whether or not they are indeed fickle. If not, then
the statement, purporting to be true, is not true, or not true in part. Now
unfortunately, the aim of using such a procedure is not, I take it, what is being
done with such statements, and it therefore misfires. First of all, you might
kill a lot more than you intend, and second, you might miss what you would
like to kill. So let me go a little bit over it again.

You would hold such a notion if you had a notion that women were
people, i.e., that the category ‘women’ is the name of some people - when no
such thing may have been claimed. The category ‘women’ is the name of a
category, and the issue is, when is it that that category is appropriately
possibly assigned to somebody? The fact that it may be correctly assigned is
irrelevant to the question of whether on some occasion it ought to be assigned,
as contrasted with something else, perhaps

The way things work is something like the following. We have our
category-bound activities, where, some activity occurring, we have a rule of
relevance which says, ‘look first to see whether the person who did it is a
member of the category to which the activity is bound.’ So that if somebody
does being fickle, or is observably being rich, you might then have a rule
which permits you to select a preferred category for seeing who they are. And
of course, using that procedure for finding the category, you may never come
across occasions for seeing that it’s ‘incorrect’ in the sense that the first
procedure I suggested would end up showing.

Now, one consequence of that procedure’s use is, if it turns out that
someone is a member of some category, then what you have is an explanation.
X is fickle. Why? Use the relevance rule. It turns out that the one who did it
is a woman. Then you can say it’s fickle because it’s a woman, and women are
fickle. One importance of these statements, then, is that they make some large
class of activities immediately understandable, needing no further explana-
tion. The statements are then to be seen as, heavily, ‘explanations.’ Where,
not only do they provide possible explanations, but, on the occurrence of some
activity, they tell you where to look first to see whether you have an

338

Part III

explanation, i.e., look to see whether the person who did it is a member of the
class or classes to which the activity is category-bound.

It is the use of such a procedure which is the important thing. You might
want to knock out this or that statement, thinking you would do away with
the trouble it makes, where what you want to knock out, if you want to do
away with the trouble, is the use of the procedure.

And the fact that it is the procedure that’s being used, provides for some
rather subtle ways of doing things like anti-Negro, anti-semitic, whatever
else, activities. For example: There’s recently been a campaign of sorts,
protesting anti-semitism by the Russian government. And the Russians
protest that they’re not doing anti-semitism at all. And the issue is, how do
they go about doing what they’re doing, and why should it be seen as
anti-semitism? Now, one of those sorts of statements we’ve been looking at
is, let’s say, “Jews are profiteers.’’ The Russians prosecute some fellow for
economic crime. One way they could do anti-semitism would be to say,
“These fellows that did these economic crimes are Jews, and Jews, of course,
do those things. They did it and it’s just what we ought to expect.” But what
they also can do, which serves perfectly well to do that task, if that’s what you
want, is the following sort of thing. They don’t claim to be prosecuting Jews
at all. They don’t mention that these people are Jews at all. They need only
- at least for internal consumption - prosecute some people and then put out
in a newspaper article, a list of the names of the persons who were convicted.
Then, given the fact that everyone knows that Jews are profiteers, ‘Jews’ now
being relevant to possibly explain who it is that did the economic crimes, one
looks to the names to see whether they are Jewish names; since Jewish names
are, among Russians, fairly identifiable, no claim at all need be made that
these are Jews for everyone who cares to see that they are Jews to see that, and
to see that what was done was something that Jews do. One has, then, an
explanation of economic crimes which can be seen as living off of anti-
semitism; otherwise the information is quite bland and unusable. And for
some purposes information is often quite bland and unusable, unless you
know how to inspect it in ways that are apparently irrelevant.

Now it happens that the fact that you have such relevance rules which
provide explanations if they apply, has further importance since it’s regularly
the case that where the question is, among competing facts, Which has
occurred? the one that has occurred is selected via the fact that it, among the
competing ones, has an explanation whereas the competing ones don’t. And
‘having an explanation’ can mean that one of those statements applies. In
some particular cases, that can be treated as quite illegitimate. So, I have an
article from the New York Times-, there was a big trial in New York where the
judge threw out a conviction in an assault and attempted rape case and
ordered a new trial for a 21-year-old Negro.

The justice pointed to the testimony of one of the jurors who said that

on more than one occasion during the deliberations remarks were made

regarding the sexual proclivities of Negroes.

Lecture 8

339

Where for some of the jurors the decision that this fellow did it could be
warranted by the fact that there was an explanation available if he did it.

That kind of thing is extremely important, for regularly people can be
puzzled how such a one who isn’t graspable via some bound category could
have done something, and it would be felt that he probably didn’t, but
somebody else did. So that if an activity occurs and the person who did it is
not known, then those bounded properties permit one to know what the
person who probably did it is going to look like, i.e., some properties of him
- and therefore, perhaps, to search for him.

And note that if that were undercut there would be some very very serious
problems. For example, if you look at the civil rights discussion and any of the
literature relevant to it, you’ll find that reports will say things like this: In the
last year, in civil-rights-related cases (not legal cases, now) where there have
been no convictions, 18 Negroes were murdered, bombed, lynched, what-
ever. Where, lacking the procedures we’re employing, that could hardly be
arrived at. All that we would know is that 18 persons had died, and there are
a variety of properties which are true about them. The sheer knowing of the
fact that this case is the case of a Negro murdered, much less a Negro
murdered for civil rights reasons, takes some of these relevance rules as
essential to the finding of what the hell is happening. And that you can then
look to some white man who is a member of a rightist organization, for
example, is not given by the fact that some X died who had certain properties,
but that one sees that it was a Negro church burned, or whatever.

And many people who would like, for example, the information about
Negroes to be undercut, to be said to be proved to be false by virtue of the
fact that it isn’t true about all Negroes, would find it very hard, then, to
preserve such matters as that in fact Negroes have been murdered in the
South, which they ought also to be able to say, and which follows as a part
of the same apparatus. One has to be careful to pick one’s weapons so that
they don’t destroy more than one would like.

Now it’s also often said, by some of my more literary friends, that when
you study things like this you’re after all studying only the things that idiots
say, or the masses, or something like that. First of all, I take it that when you
hear, for example, of a civil rights murder, you perfectly well see that it’s a
civil rights murder, even if, when you hear somebody talking about Negroes’
sexual proclivities you may say, “Well that’s not true’’, and the same for any
of the other substitutions. In fact, I take it, no matter how sophisticated you
are, you use such procedures. And they’re quite important to use. And, as I
mentioned earlier, they’re important in other ways than dealing with bad
information, i.e., it’s important to know that lawyers are rich so as to permit
somebody who wants to be rich to know what to do with their lives.

In any event, if one likes such little games, one can collect, as I, in moments
of harrassment have collected, statements of some of the most eminent of
living or dead geniuses, which perfectly well satisfy the analyses we’ve
developed. Because one really ought to see that what we’re dealing with is not
something that’s restricted to the uneducated and that people with a college

340

Part III

education are freed from such prejudices as they are freed from using
stereotyped information. Here is just one of the statements I’ve collected. It’s
a piece of testimony.

Questioner :
Witness :
Questioner :

Witness :

Questioner :
Witness :

Questioner 2 :
Witness :

Questioner :
Witness :

Were you acquainted with Fuchs?

Yes, I knew Fuchs quite well.

Did you have any reason to suspect his integrity or
dependability or whatever was involved in the subsequent
disclosure?

Not particularly. He was a rather queer person, but then
under those conditions queer persons occur. I did not
suspect him particularly. He was clearly not an ordinary
person.

What I’m getting at is whether you had reason to believe
he was a Communist.

I think I did not know about him, no. I did not know
about him. That he was a Communist, practically until the
whole affair broke.

Practically what?

Until it became known. Until he confessed. Or rather,
until he was shown.

At the time you learned about it were you surprised?
Look. I was not surprised in this sense. That he clearly was
a peculiar person. So if it turns out about an ordinary run
of the mill person that he’s a conspirator and spy, you are
shocked and surprised. He was a very peculiar persons,
with respect to whom I didn’t have much experience. Of
course I was surprised by the fact that there had been such
a thing, that a spy had been so well placed.

That is to say, if you’re told something and you know somebody’s queer, then
anything much they do is understandable. The piece of testimony is from the
Oppenheimer hearings and the witness is John Von Neumann; surely
someone who goes as among the great giants of our day, intellectually.

If one wants to propose that what I’m saying only holds for 1 6-year-old
kids or people with less than a highschool education who read Life and Time,
then I suppose it’s only a misunderstanding. That is, that you are supposing
that what I’m saying is that such information as “women are fickle,” “queer
people do anything,” and things like that, have a character which I am not
at all proposing that they have. And one also ought to get some idea about
what reforms it would take, what you would want to do; which is perhaps,
for example, to undercut the relevance rules, to say that if you hear certain
activities done, don’t look to category X, and things like that. It might pay
to examine what else would happen once you changed one of them, and
consider whether you would like to change it.

Lecture 9

Lecture 9 consisted of two distinct parts; the first part, pp. 1—8, dealt with
matters carried over from lecture 8. The second part, pp. 8—12, started a
discussion that was carried over into lecture 1 0. In this volume the appropriate
parts have been absorbed into lectures 8 and 10, and there is no lecture 9-

341

Lecture 10

Pro-verbs; Performatives ; Position
markers ; Warnings

I’m going to begin to deal with this statement from the data:

Ken : I still say though that if you take if you take uh a big fancy car out

on the road and you’re hotroddin’ around you’re- you’re bound to
get you’re bound to get caught, and you’re bound to get shafted.

It’s a tremendously complicated thing and I’m not going to give it what it
requires. I’ll begin off with some little pieces.

One of the questions we could ask concerns what it is that that first bit, “I
still say though that ...” is doing there. Is it to be seen as simply redundant,
in the sense that the statement could perfectly well do without it - obviously
he’s the one saying it? Where one might take it that the statement might have
been useful if he were talking about a past event and introducing a quotation
of something he happened to have said earlier. Given, however, that it’s a
present occurrence, what is it that - focussing only on ‘I say’ for the moment
- does, if anything.

I’m going to have to first introduce an unfamiliar term which, however, I
haven’t invented. The term is ‘pro-verb.’ Following is the statement in which
I found reference to at least the existence of such a thing. It’s from an article
called ‘On the semantic structure of language,’ by Uriel Wereich in Joseph
Greenberg (ed . ) , Universals of Language (1963) page 126.

All languages have ‘pro-forms’ such as ‘he’ which substitute for other
forms to avoid their repetition within a unit of discourse considered as
‘the same.’ But pro-forms are on the whole very unevenly distributed
with respect to the parts of speech. Perhaps all languages have
pronouns, but few have pro-verbs. English is perhaps unique among
European languages in having, in ‘do,’ at least the rudiments of a
verb-phrase substitute.

That is to say, roughly, that ‘do’ works something like the same way that
more familiar pro-terms, i.e., pronouns, work. For example, in the uses that
his notion restricts itself to, one might say, for the pronoun, “John put on his
coat and he put on his hat,” ‘he’ substituting for ‘John’ in the second clause.
Alternatively, for ‘do’ you might say, “Mary baked a cake and so did Jane;”

Includes pp. 8-12 of lecture 9, Spring 1966.

342

Lecture 10 343

‘do’ doing similar work, replacing ‘baked a cake’ as ‘he’ replaced ‘John’ in the
other.

The term wasn’t invented by Weinreich, and I don’t know who did, and
can’t find out easily. It’s clear in our data that ‘do’ works that way. And if
you’re interested in the historical development of this usage there’s an article
called ‘The auxiliary “do”: the establishment and regulation of its use in
English’ (which ends about the 14th century for this guy), by Alan Ellegard,
in Gothenberg Studies in English, volume 2 (Stockholm, 1963).

What I want to do is develop what looks to me like another pro-verb. It’s
not as extended in its use as ‘do,’ and that is, ‘say.’ And the question is, first
off, what is the class of things for which ‘say’ serves as a pro-verb? The class
of terms that ‘say’ seems to serve as a pro-term for are at least roughly
collected by Austin’s notion of performatives.’ There are two books, and
they’re books you ought to read. In his Philosophical Papers , chapter 10,
there’s a paper called ‘Performative utterances.’ Then there is a whole book
about these things called How To Do Things With Words, ]. L. Austin
(Harvard University Press, 1962). They are things like, ‘I promise,’ ‘I offer,’
‘I bet,’ ‘I assume,’ etc.; various activities that you can do in that form. That
is, roughly, where something that you say is a formula for doing some
activity; it’s done in the saying. You can make a bet by saying ‘I bet,’ a
promise by saying ‘I promise.’

And it’s extremely important that one restrict oneself to the first-person
present form, i.e., one doesn’t take all its tense and person possibilities. ‘I bet’
is altogether different than ‘He bets.’ ‘I bet’ can do betting. ‘He bets’ reports
on something somebody is doing. So, ‘performatives’ are all first-person
present. And in that way, ‘I say’ is very different than ‘I said.’ ‘I said’ can be
introducing a quotation. Let me now consider what ‘I say’ can be doing.

What it seems is that for some of those performative verbs, instead of
doing the activity with, e.g., ‘I claim,’ ‘I assert,’ etc., one can do it with ‘I say,’
where the hearer will perhaps put in the proper performative verb (if that’s
the way hearers do things). What one may then have is a weak, or a generic
type of way of doing any of these things without specifying the one, though
the one is ostensibly clear from the rest of what you’re doing. That is, for
example in this case, it being a ‘discussion,’ and one of the ways one
participates in a discussion being to ‘take a position,’ one can say ‘I say’ when
what one is doing is - if you want to paraphrase it - saying in effect, ‘My
position is,’ ‘I assert that.’

Now, the difference between ‘merely saying’ something, and doing what
can be done with ‘1 say’ are, among other things, the following. First, you’re
often making some commitment, in the case of a discussion perhaps, to
defending yourself. Or to maintaining that argument. Or to answering
questions if it’s unclear. Or to living up to it. And, in that it involves a
commitment, it may be very important who you’re speaking on behalf of;
that is, who is it that you’re representing. “In the name of the United States
I hereby claim this piece of territory.” For any such thing it is an important
matter who you’re on behalf of, because if you’re not properly on behalf of

344

Part III

something, then you can’t do the thing you’re trying to do. If you’re not an
official who is deputed to doing marriages, for example, you can say “I marry
you” all day and you don’t do it.

And in our segment, we just previously had a statement of the form ‘We
were,’ i.e., “We were in an automobile discussion.” And now T can be, not
merely in contrast to that, but the thing is being asserted on ‘my’ behalf, and
that may be all it’s asserted on behalf of. So it’s when we get the character of
‘say.’ and what we could call the ‘generic performative,’ or a ‘pro-term for
performatives,’ that we find that the initial pronoun, or initial affiliation,
becomes quite relevant. For one, we might want to know where to go to
complain. Who to address if we disagree.

Further, while if he were saying ‘I said,’ then you might say, ‘No, you said
something else.’ But for ‘I say,’ if you wanted to point out he was wrong, you
wouldn’t propose that he doesn’t say that; what you would want would be to
say that the argument that he’s committed to is wrong.

Now it’s also the case that ‘said’ can have this usage. So, for example,
regularly with kids, if somebody gives permission to do something, e.g., “Can
I watch television?” “Yes,” then comes back ten minutes later and says,
“Turn the television set off,” the kid says, “But you said!” Where they’re not
simply saying ‘I’m reporting a quote of yours,’ but ‘You did something which
involved making a commitment (one of this class which can be referred to by
the term “say”) and what I’m doing is invoking your commitment.’

So: So far I’ve been engaged in proposing that with ‘I say’ what’s being
done in part is to mark that the fellow who has started off his statement with
“I still say though ...” is making a position in an argument, where the first
term, ‘I,’ has its relevance by virtue of the fact that for the class of actions
which ‘asserting’ or ‘making positions’ involves, there’s a question about on
behalf of whom it’s done. That question can be handled in various ways, one
of them being that after some categorical term is used, like ‘we,’ it could be
claimed, ‘What’s your entitlement to act on their behalf?’ ‘I’ is a trivial case
in this circumstance, since for the taking of a position, one is the best
representative of ‘I.’ So, with ‘I,’ you solve the problem of asserting on whose
behalf, and in using ‘I,’ have that warrant; whereas for other terms, the issue
of represen tivity can be raised and is regularly raised.

Now we can get to ‘still’ and ‘though.’ These are general position markers
for some type of activity within which the given activity is done. So if ‘I say’
is an activity done as part of some argument, or discussion that can become
or is becoming an argument, then ‘still’ and ‘though’ are ‘positional markers’
within a discussion or argument. Their character is to be found via the sort of
general activity that’s being done, i.e., the sort of position they mark can vary
with each activity. They needn’t be done in combination, and in combination
(i.e., as ‘still . . . though’) they can mark a different position than they do
singularly (i.e., than either ‘still’ does or ‘though’ does). And they’re not the
only positional markers, either.

What sort of positional markers they are is something that can be seen as
rough, but may well be as exact as one gets, or as one needs. In combination,

Lecture 10

345

‘still’ and ‘though’ seem to mark the following. Whatever particular activity
is being done within this general activity is at least a third activity of which
some prior one was done by the same person. In the case of an argument,
then, it’s not a second statement. ‘Though’ alone can be a way of marking a
second statement in an argument where the first is made by somebody else.
And by ‘second,’ I don’t necessarily mean this is his second utterance. 1 mean
only that there’s some prior statement by somebody else, where this one can
be the first by the one who’s speaking. So, for example, at some point in an
argument where some person hasn’t yet spoken, their first utterance may be
introduced with something like, “You know what / think, though.’’ Where,
now, ‘still’ in combination with ‘though’ seems to serve to inform (if it’s
relevant to inform, and it’s relevant in serious ways in this case) that one of
those preceding positions was taken by the speaker and he’s going to reaffirm
it.

There are a variety of relevancies for using these markers in an argument.
In the case at hand of course, we have a new person on the scene who hasn’t
heard the arguments before, and indeed, doesn’t know that the argument is
going to begin up again, but they might now go on to talking about anything
else. With that beginning clause he’s informed that they’re back in the
argument and that the position being taken by the person who is now going
back into the argument is to be seen as something reopening it or continuing
it and not starting something new. And it is to be seen as a reaffirmation, but
not just a reaffirmation but a reaffirmation in the face of some opposition. And
if the argument itself is familiar enough, it could routinely be available to the
new entrant, on hearing just this fellow’s statement plus his argument, that
there were opposing arguments, and what the opposing arguments were.

There are other things involved in the use of these positional markers. One
of them is that they involve a way of adding another activity to the given
activity of making an argument. If he simply makes his argument again,
that’s one sort of thing, about which it could be said, e.g., that he’s being,
whether he knows it or not, stubborn or dense. But by introducing the
positional markers which claim ‘reaffirmation in the face of opposition,’ he’s
providing for the fact that he does see what he’s doing as ‘stubbornly
maintaining his position.’ So, one way one can do being stubborn is the use
of that combination of position markers. Of course they could be used
wrongly in a variety of ways. For one, there may be no opposition. Intervening
arguments may have been intended as supportive, or may be very weak, such
that to come on being stubborn can make you look foolish (or it could be cute,
e.g., if an infant thrashes around in a wading pool as though they’re possibly
drowning). But these things do seem to involve, then, a claim of the
reaffirmation being a ‘stubborn reaffirmation.’

So the phrase ‘I still say though’ does a whole set of tasks: It reintroduces
the discussion as an argument, gives a great deal of information - or can — to
the person who wasn’t there before, and provides, at least for those who were
there, that the position that was taken before is being stubbornly held to.

These positional markers are, then, one way that the character and the

346

Part III

relative where-in-it one is, in some kind of activity like an argument, is
persistently affirmed. And indeed, the character of a discussion as ‘an
argument’ is in part maintained via the use of these things.

Let me turn now to the long clause that follows “I still say though that,”
i.e., the statement, ‘‘if you’re hotrodding you’re bound to get caught,” and
give it an initial characterization. Roughly, you could consider it for now as
a ‘warning.’ Now warnings are extraordinarily tricky matters. The format one
uses for a warning can, with no change in what you say essentially, be used to
do as well, alternatively or in combination, any of the following, at least:
Threatening, predicting, advising, challenging, daring, promising. We’ll have
to work at how that can be so, and what the consequences of that are.

If you look at the next statement:

Al : Not unless ya do it right. That’s the challenge- that’s the challenge

you wanna try and do it right so you don’t get caught!

you’ll see that if I’m correct that this was proposed as a ‘warning’ it was heard
as a ‘challenge.’ And let’s try then to do just that much, i.e., to show how a
warning can stand as a challenge. Something like this seems to be involved:
One of the ways that a warning is in part successful is if it brings the warned
person to recognize that the activity that he’s proposing to do, seems to be
about to do, might do, is an instance of the class ‘dangers.’ But if you succeed
in doing that, then that alone done, you’ve partially provided for the failure
of the rest of your task - which is perhaps to bring them not to do it - by
virtue of the consequences of something being a case of the class ‘dangers.’ For
that class, quite independently of the issue of warnings, has a series of
properties are that important.

The character of ‘dangers’ is, for one, that there are certain memberships
that can only be achieved if one does something which is a case of the class
‘dangers.’ In its narrowest scope, the only way to be a hero is to do something
dangerous. There are other membership classes which are larger than ‘hero,’
for which, while it may not be the only way, it’s a way to get membership in
them; especially if, say, one is close to, but not in, membership, and one has
doubts about one’s membership and others have doubts about it.

That’s set up independently of warnings, but can mesh with warnings; i.e.,
the way you have something guaranteed as being a member of the class
‘dangers’ in the first place can be that somebody warns you not to do it. The
warner tells you that at least for them as an audience, that activity is
dangerous. They may warn you to get you not to do it. But if for you it’s an
issue that if something is proposed as dangerous then it’s an eligible way to
do those things which you can accomplish via doing dangerous things, then
that provides the warrant for doing it.

Further, I think it is regularly if not always the case about categories for
which membership is controlled by doing something dangerous, that they are
from devices whose categories are positioned. So the category one achieves
membership in via doing something dangerous, is the highest in the device;

Lecture 10

347

where ‘hero/coward’ is an obvious example. Or, for example, ‘men/boys’ is
ordered in just the same way: Doing dangerous things is what differentiates
the men from the boys. And the character of ‘dangerous things,’ as providing
access to the higher of positioned categories in any category which is bound to
a dangerous thing is, I think, as good a criterion as you want for the fact that
they are positioned. That procedure is not specific to this culture at all; it’s
quite general. Built into many ‘native’ cultures is just such a test, which
consists of some way that membership in some higher-positioned category is
proved via having done something which is, by definition, or recognized to
be, dangerous. And the same holds here. So that we can see, then, what kind
of link there is between offering a warning and having it heard as a challenge.

With that, there are some familiar ideas, which I for one don’t readily
understand, which I think can become clarified. One is the notion of
‘ambivalence.’ One can see, for example, how it is that warning somebody is
something one can do via a combination of love and hate. Seeking to protect
them and/or to drive them to something dangerous.

At this point we can come to see the systematic sociological basis for a
notion like ‘ambivalence’ holding for an activity like ‘warning,’ given that
warning is locked into these other things in a way that is perhaps not separable
from them, and these things are quite opposite in their character. If, for
example, there is something which, if you want to be a man you have to do,
or if you have doubts about whether you are a man you might do to show it,
then one can see what the strains are that are involved when a mother says,
“For your own good, don’t do it.’’ And one can see the extent to which her
warnings, which are of course seen by her as purely derived from her love, are
ways of preventing maturity, maintaining the dependence, etc., of the object
she warns - which, of course, if it refuses the warning, accepts the challenge,
becomes a different object: No longer a boy, a man. As such, no longer
subjectable to the same warnings. And that’s of course a fact about dangerous
things; they are differentiated. Things that are dangerous for one, are not for
the other.

The motives of warners are, then, presumably always suspect. I don’t mean
that you have to be suspicious of them. What I mean is that the motives can’t
be clarified from the fact that someone has done a warning, in some exclusive
way, e.g., if they did it then they must love you. Or if they did it then they
must hate you. But they both can be present.

Lecture 11

‘1 You’

I’m going to talk about ‘you’ in this sentence we’re examining, but first I want
to make a brief remark about another aspect of it, a phenomenon which
recurs in the speech of this fellow Ken. There are two of them present here:
“If you take if you take uh a big fancy car. . .’’ and “you’re you’re bound to
get you’re bound to get caught.

Ken : I still say though that if you take if you take uh a big fancy car out

on the road and you’re hotroddin’ around, you’re- you’re bound to
get you’re bound to get caught and you’re bound to get shafted.

These things have been located; the term that’s been applied to them is
‘buildups;’ they’re very characteristic of rather young children, say, two years
old. There is a brief discussion of them in a book called hanguage in the Crib
by Ruth Hirsh Weir, pages 81-82. Presumably for young children they occur
by virtue of the fact that the ability to build a sentence is quite limited, and
you get a steady expansion from some small part to some larger part which
is syntactically coherent with it. I suppose that would provide a rather
technical way of formulating somebody’s infantilism, if one cared to. (The
Weir book is fairly new, but it’s been enormously well received, and all the
data that’s used in the analysis she develops, is presented, phonetically
transcribed.)

Now the discussion of ‘you’ that I want to work at involves dealing with,
in its general character, a problem to which little attention has yet been paid,
though it’s quite clear, even for linguistics, that it’s probably going to be a
very important problem. The only relevant paper I know of for linguistics is
‘A grammar for the hearer’ by Charles Hockett by in Roman Jakobsen (ed.),
Proceedings of Symposia in Applied Mathematics, vol. 12, Language in its
Mathematical Aspects. It’s on page 220, I think.

The paper itself is completely theoretical, but it suggests the kind of
question I want to address, which is in part: Are there some sequential
procedures used to detect which of possible alternative meanings for a word
might be intended by its producer; which concerns, then: Is there some
sequence in which a sentence is parsed, or come to be understood? Do
Members wait until the end, do they stop in the middle, do they try to detect
it all over its course? Clearly enough from what we’ve seen so far, there are
plenty of places where people do not wait for the end, since there are some
places at least where they interrupt and complete an utterance. To do that is
something, presumably at least, approximately similar to the way that in relay

348

Lecture 1 1

349

races, when somebody’s about to go, he starts running before he gets the stick,
and runs along for a while with the other. And one would imagine that the
detection of sentences which are completed by some second person involves
him in moving along to fit his remark in at some such point, and not waiting
until it’s completed or until the last word, to fit his own remark in. Something
of this sort occurs in one of the therapy sessions where, were the second speaker
to wait until completion of the first speaker’s utterance, his own remark would
not come off. They’ve been talking for a while about reincarnation:

Ken : I was- I was on a road on the way to Rozwell New Mexico. An’ I

was, y’know, plonkin’ along at a regular speed,

Roger : An’ a grasshopper jumped onto the road who you recognized!

The issue of a possible sequence in which something comes to be
understood, arises in part because of the character of ‘you.’ And I will consider
some of the properties of that term now. In English, ‘you’ is in the first
instance systematically ambiguous, in that it does not discriminate between
singular and plural reference. And in that ambiguity, one has the ‘this-and-
that’ format - a way of building richness into a compact form. That is, you
get words which mean, not ‘either-or’ (where either you mean this or you
mean that, and they don’t have anything much to do with each other, but it’s
not clear which you mean on some given utterance) but, whatever meanings
the word has, they may be carried. So that ‘you’ may mean ‘you alone;’ if it
does not, then it means ‘you and others;’ but the ambiguity does not involve
deciding whether ‘you’ or ‘others’ was meant.

Further, ‘you’ is the proper, normal way to address some other, i.e., in
contrast to supposing that it’s a substitute for a name, which it isn’t. Now that
may then serve to set up the following procedural sorts of possibilities. The
alternatives are singular and plural, and perhaps there’s some sequence in
which those alternatives are gone through. For example, perhaps the hearer
who hears ‘you,’ hears first ‘me.’ And if that doesn’t work for some reason,
he then goes to the next step which is ‘me plus somebody else (or, some
others).’ That is to say, the second step in the procedure is an expansion, and
not, e.g., an exclusion of ‘me.’

That openness of ‘you’ means that ‘you’ can in fact be a way of talking
about ‘everybody’ - and indeed, incidentally, of ‘me.’ For ‘you’ stands as a
pronoun for the set of terms: ‘everyone,’ ‘someone,’ ‘people,’ etc. Here is
some relevant data:

A: Why do you want to kill yourself?

B: For the same reason everybody does.

A: What is that?

B: Well, you just want to know if someone cares.

J : I’m curious. I’d like to know what happens after someone dies.

R\ You lay in the ground. They have a saying, you’re planted.

Part 111

350

A: You have a pretty strict conscience, haven’t you? A pretty strict sense

of what’s right.

B: Well, everyone does.

A: No, not true. You don’t think so, really. A lot of people sort of slip

through life just doing what’s easy.

B : Well, you have to think of other people, too. You have to think of

their feelings.

I think it’s pretty clear in these cases that ‘you’ stands as a pronoun for
‘everyone.’

And then we can see a difference between the plurals ‘we’/‘they’ and the
plural ‘you,’ which would account for the use of ‘you as the pronoun for
‘everyone,’ perhaps. That is, the expansion from singular to plural ‘you’
permits an indefinitely large expansion, whereas either of the other two plural
pronouns excludes the other. ‘We’ excludes ‘they’ and ‘they’ excludes ‘we.’
‘You’ as it expands and eventually meets ‘everyone,’ excludes no one.

It comes to mean ‘me,’ indeed, also. And in the data above, it clearly
means, in the first instance, ‘me.’ But you come at ‘me’ in an especially strong
way. That is, you come at it by virtue of the fact that the hearer hears first
‘you,’ himself, and he knows it it’s not only that, or maybe he doesn’t think
it’s even that, but his next step is not, given the way ‘you’ can be used, an
exclusion step (unless he says it’s wrong), but an expansion. If he wants the
things to include ‘you,’ the speaker, in that expansion, then he gets to see
‘you’ as standing for ‘everyone,’ which then includes the speaker.

There are very good reasons for a speaker doing that sort of thing. A first
is, of course, if you’re asked some sort of question which you know in the first
instance is one which has the sort of answers that you must defend when
you’ve given them, and you’d like to have an answer that has a built-in
defense for it, then one of the defenses would be to show that there’s nothing
at all problematic about your answer. For “Why do you want to kill
yourself?’’ almost any reason can then be attached - proposing you’re ill, or
whatever else. If you arrive at it via this statement being true for everyone, at
least perhaps everyone who is suicidal, then there’s nothing special about you.

What’s especially important to see is that if a speaker knows the procedures
the hearer uses to arrive at some way of deciding the reference, and what the
outcomes of those are, then he can of course drop in a term whose use he
knows, so as to get the hearing that he wants.

Now this last point can be dealt with, with related though somewhat
independent materials. In our segment of the group therapy session (third
page of the transcript) we find a very common kind of thing:

Roger : Ken, face it, you’re a poor little rich kid.

Ken : Yes, Mommy. Thank you.

One could imagine that to “Yes, Mommy. Thank you” everyone turns
around to find out who’s being addressed. Clearly none of them, and certainly

Lecture 1 1

351

not the person who spoke, is the mommy of this fellow. How is it that the
person who is called “Yes, Mommy’’ knows perfectly well that he’s being
called “Yes, Mommy’’? How would we go about coming to show how that’s
so - again, in such a way as to see how the doer of that will not be seen as
simply making an error. And indeed, if he were seen as making an error that
would be interesting because if it were seen that he were making an error, then
there would be at least the awareness that the name was not an appropriate
one for someone who might expectably be being addressed.

There’s a rather simple way of coming to see how that works. First of all,
there are places within statements where terms of address can go. And given
some of the paired statements (question-answer, for example), if you have a
second of some pair, then to whom that second ought to be directed is known
in the first instance, i.e., the answer is expectably directed to the one who
asked the question.

Assuming for now that “Ken face it you’re a poor little rich kid” is an
insult, the fellow who did the insult is the one to whom some retort would
be addressed. Therefore, whatever term may occur in what’s to be heard as
a retort, is in the first instance to be heard as addressed to him. And then
the issue is, how is it that it characterized him, and not , given a term, find
who it is that it might address, or if you can’t find anybody treat it as an
error.

The assessment of it as an insult-return can be made by the recipient; he can
then take the given term and know that it’s addressed to him, and know that
everybody else knows that it’s addressed to him, and also know that the
person who used it used it knowing that he would hear it as addressing
him. Which all can be done by virtue of the fact that the sequence in which
that thing will be picked up is available to all of them - as in the case of
‘you.’

One of the things that ought to be noted is that some sequence in which
a hearer goes about doing the job of hearing is absolutely crucial to coming to
see what it is that these people are doing with their statements and with the
receipt of their statements.

Now it’s also of course the case that ‘we’ has the same difficulties in its
various uses. But there’s at least a tactical difference between saying ‘we’ with
the plural intention and ‘you’ with the plural intention. The difference is, if
I say ‘we,’ that has some plural reference automatically. However, it does not
automatically have a plural reference which includes the person being spoken
to. It can be exclusive of the recipient. “We were in an automobile discussion”
is ‘we’ in exclusion of the entrant. This possibility provides for a way of saying
‘everyone,’ which is: If someone who needn’t include themselves in some class
in which the recipient is a member - or someone else that the recipient is
talking of is a member - wants to say in a stronger way than ‘you’ (which can
be misheard as only the recipient) that ‘everyone is that way,’ they can say
‘we.’ For example, a woman is talking about her husband, who responded to
an appeal of hers in something she thought was a rather hard-hearted way.
And the psychiatrist is wanting to suggest that her appeal may have been not

352

Part III

clear enough, where she wants to suggest that her husband’s response is
evidence that he doesn’t care.

C: Wouldn’t you think that if a person found someone that tried to

commit suicide they would realize that it was serious.

S : Sometimes people do, and sometimes they don’t. Sometimes they

need someone to tell them straight out. We can be very blind to the
things around us.

That is, by virtue of the sequence in which ‘we’ is heard, the inclusion of the
speaker is partially gratuitous; is heard as doing something he ‘needn’t have
done.’ Where, if you use ‘you,’ it in the first instance includes the recipient,
providing for, in the above case, her hearing ‘me,’ which may be sufficient
reference for the remark tied to it, and so not providing for the further step
which would include the speaker. It is on the recipient’s option or the
speaker’s intention - insofar as those coincide - that ‘you’ can refer to some
category which includes ‘everybody else.’

Furthermore, it is regularly the case that when I choose ‘you,’ speaking to
a singular person I’m talking to, ‘you’ is not the singular of the plural ‘you,’
but the local representative of somebody I would otherwise refer to as ‘they.’
For example, one might say ‘you people,’ talking to some person or persons,
who, were they absent, you would call them ‘they;’ e.g., ‘You Negroes.’
Where the issue of the use of ‘they’ is given via, say, the fact that for some
device, I’m in one category and you’re in another. My category, when I refer
to it without a name, I call ‘we,’ and yours I call ‘they.’ When you’re present,
I say ‘you.’ Now in saying ‘you,’ I have that as an alternative to saying ‘we,’
in that the two of us present can regularly be referred to as ‘we.’ So what we
get, then, is where I could say ‘you,’ I could also regularly say ‘we,’ and when
I say ‘you’ in that circumstance, I will regularly be using ‘you’ as a local for
‘they.’

One way to see that usage being understood, is that if I say to you in
conversation, ‘you,’ you may well come back to me saying ‘we,’ exclusively
of me. That is, “What do you think?” “Well, we think. . .,’ i.e., you
understand that my use of ‘you’ meant ‘you as a member of some category
of which I’m not a member.’ And you then come back, qua member of that
category, as a representative of it, rather than that ‘we’ now being inclusive of
me. It can announce categorial membership being used. Since these uses are
perfectly obvious and simple, you would have to suppose that a Member who
produces that statement, and one who has to deal with it, also have some
apparatus for deciding what’s being done with ‘we’ and what it refers to.

There is a further point I want to make about pronouns, and we can get
to it through a quotation I’ll read you from Language in the Crib by Ruth
Weir, pages 73-74:

Of the personal pronouns the form ‘it’ has by far the widest occurrence,
over 100, and is firmly established in the child’s speech pattern.

Lecture 11

353

Occasionally it is over-used as a substitute for a plural noun, at the
expense of ‘them.’ The frequent use of ‘it’ confirms the child’s
classification of nouns into personalized and non-personalized classes as
was pointed out. The non-personalized nouns permit substitution by the
pronoun ‘it’ rather frequently, whereas the personalized nouns, includ-
ing names of persons, favorite toys and animals, rarely are substituted
for by pronouns.

The point I want to make is that I do not agree with talk of pronouns as
‘substitutes for nouns.’ In some ways, at least under certain very natural senses
of ‘substitutes,’ that is obviously wrong. And as for others, it’s a big issue.

If we treat the corpus of pronouns as they are conventionally treated,
including terms like, e.g., T and ‘you’ as pronouns, then in the usual
environment of those terms it’s an absurdity to say that they are substitutes
for nouns, in the sense that the proper term to use in discourse when one
intends to refer to oneself is T or things like it, and not - except under very
special circumstances - one’s name - which would be seen as being used as a
substitute for the pronoun. Now we could say that ‘I’ doesn’t belong among
the pronouns in the sense that were I to talk of the blackboard as ‘it,’ ‘it’ being
a substitute for ‘the blackboard,’ that ‘I’ is not such a term. In the rules of
conversation terms like T and ‘you’ are the normal, proper terms and are not
substitutes for some other terms.

We could say that pronouns serve as ‘reference preservers’ in the sense that
they tie together some current piece of talk to some past piece of talk, and in
some cases, to some future piece of talk. And they do that in a sense that the
talk wouldn’t be better if it were rewritten with the nouns that they ‘substitute
for. ’ It would often be worse. And furthermore, learning to talk involves that
you learn how to put the pronouns in. And you can’t, I think, sustain talk
without them.

Lecture 12

Warnings, challenges, and
corrections; Explanations ;
Complaining— Praising; Games

I want to talk a bit more, now, about ‘warnings and related matters. One of
the things one can perhaps dig out of materials such as the kind we’ve got
(but I take it you’d need a considerable amount to do it) turns on the
following sorts of things. The fellow makes a warning, and with it he might
perhaps be engaged in threatening the other - if he does such things. And we
might look to see what sort of apparatus he invokes in establishing his
warning. Warnings can be done in dramatically different ways. I could say,
“If you step across this line I’ll punch you in the face,” and that sort of
warning might be seen as rather different than those which go, “If you hit me
again I’ll tell my father,” and whatever further invocation of more or less
massed power might be involved in getting up sufficient courage, perhaps, to
warn someone. And one would like to see whether there is some systematic
difference in the sorts of machinery invoked by various people doing
warnings.

So, for example, there are two persons to this hotrodding warning, at least.
But the person who hears it as a challenge is one who does warnings
throughout this session - not in our segment but elsewhere, and he warns this
very fellow here. But when he warns him, he warns him as a personal threat:
“I will bang you around if you do something.” Whereas, again, this fellow
makes another warning, this time to the entrant, on the sixth page of the
transcript, as the other two are preparing to ‘judge the entrant' s character.’ He
says, “You’re gonna get killed” to which he adds, “Keep your mouth shut
and don’t say a damn word.” Now again, the warning is not involving
himself as the enforcer. And what exactly you might be able to say about
people being systematically differentiated by the apparatuses they use to make
warnings, I’m not sure, but it may well be that you could. Whatever case one
wants to make, of course, needs be made from some data.

In that line of consideration, shifting matters a little bit, you might also
consider the terms in which the external world is invoked in some speech. One
might go through these materials to see, when people other than these
participants, or when social organization, things like that, are referred to, how
they are referred to. Is it regularly via warnings? If so, one has some of the

Starts with the last three pages of the original lecture 1 1 .

354

Lecture 12 355

sense that might be involved in saying that neurotics or psychotics see the
outside world as threatening.

Whether either task is doable, I don’t know, and I take it it would involve
some larger corpus of materials, since you’re making basically a statistical-
type argument. But in any event, we oughtn’t to stop by saying only that he
made a warning and that he made it in a way that one can make warnings,
but we ought to note a variety of other possibilities- that are involved. For
example, that wanting to make a warning, it may be the case that the best
warning you can come up with is one which involves no less than the police
power. Or, for example, that the ways for this fellow that the external world
is relevant is as enforcers.

Some of the things that are present in this sentence I’m going to leave for
now, and consider them when we get to this fellow’s next speech, that long
statement starting off with “In that Bonneville of mine.’’ But let me just offer
one for consideration. You might notice that there isn’t any explicit indication
that a crime was in fact committed. If you go hotrodding, you may surely
commit a crime, but you don’t only commit crimes, and of course most of the
things that you can do while you’re hotrodding are legal. But it seems quite
obvious that the thing is presented, e.g., with the term ‘getting caught,’ as
indicating that the arrest involved some illegal act.

And furthermore, the adequacy of the account that he gives is not undercut
by any of the other participants, by, for example, suggesting that while you
may indeed get caught, probably it wouldn’t have been because you were
doing something illegal. That is to say, the others don’t see this fellow’s
explanation as inadequate. Rather does it seem that if you were doing
something illegal and you got arrested, then not only is it the case that you got
arrested because you were doing something illegal, but nothing much more
need be said about the arrest. That can be contrasted to an earlier statement
where they consider getting arrested without doing something obviously
illegal, and use, to refer to the policeman who does it, that he “wants to be
mean.’’ Now that’s something which is exceptionally interesting, and really
worth giving a lot of thought to. I’ll put the matter very generally or
abstractly because at this point I can’t present it in the kind of detail I would
like.

It seems to be the case that types of explanations are ordered into layers, or
domains, as one may please to call them. Some of these we would roughly call
‘Members’ sociological accounts;’ others ‘Members’ psychological accounts,’
etc., whatever the others may be. And it seems that there is some kind of
substitutability operating within a domain. If a domain is relevant to an
account, then there can be a strong version of the economy rule holding. And
the strong version means that there can be an explanation which is the
explanation, with no more needed or proper. Such that, for example, if the
cop picks you up when you’re doing something wrong, he picks you up
because you’re doing something wrong, and anything else added as an
explanation is something that’s not right to do. That is to say, it’s not right
to add such an explanation as, “He did it because it’s the end of the month

356

Part III

and they have to keep up with their quota,” or . .because he was nasty” or
. .because he didn’t like us.”

Given the economy rule and given the domains, there may well be
exclusivity between domains, such that if some account from one domain is
relevant, accounts from other domains are excluded. That is, it’s not that you
need one account from each, or several from each. If you have one from a
relevant one (or whatever number necessary from a relevant one), then you
may well exclude any from the others.

Those separations or those rules are enforced, and the penalties for violation
of them are very very serious. Let’s say, for example, you talk in the following
kind of way and you see the world in the following kind of way:

A: Why would you get fired?

B\ Why would I get fired? Because people are bastards. That’s the best
answer I can give you.

A: You mean it’s some trouble between someone else and you?

B: No. It’s trouble because of the nature of dishonest, corrupt, immoral,

indecent people. People are interested only in themselves and the hell
with the rest of the world. In other words unless they can make a
dollar out of it they won’t do a damn thing in the world.

If you talk like that, given questions like, not “Why did you get fired from
this job?” but “Why do you get fired?” then you’re going to be heard as
‘paranoid.’ And that is a rather serious sanction controlling such movements
between layers.

Now the kinds of exclusion between layers, levels, domains can be more or
less supported by various kinds of institutions, such that courts will only hear
certain kinds of explanations as appropriate, and the like. And one can
perhaps lay out that the kinds of attacks that are formulated about Freudian
psychotherapy by Marxists consist of a pointing to the fact that some domain
(the emotions) is used exclusively, which is otherwise used by Members, and
also a very special unit of social organization (the family), and any other
accounts of troubles are excluded. Where what the opposition proposes is that
explanations for things like why X feels lousy, why he can’t do various things,
are to be formulated by reference to, for example, his social position, the
structure of society in the sense of its economic structure, and the like. Let me
bring this home to you with among my favorite of all quotations. It’s so
striking that I think very few people would propose it, but in principle it’s the
sort of thing that everybody does all the time. It’s from the Washington Post,
a kind of social welfare article:

Elizabeth was a bright, attractive 1 4-year-old who had grown up in a
house crowded with in-laws, brothers, sisters, nieces and nephews. Her
parents both worked and were rarely at home. Her older brothers and
sisters also worked, leaving the youngsters to manage for themselves.
No one particularly cared whether Elizabeth or any of the other children

Lecture 12

357

got to school, needed anything, or were happy. They were just left there

to grow up as best they could.

As a result of these conditions, Elizabeth became pregnant.

Such attempts can be revolutionary, not in the sense that they want to
change the structure of the society, but they want to break some of the
separations and some of the rules of relevance that decide what an explanation
would be. You could perfectly well have major changes in the way an
economy is organized without much affecting, perhaps, what it is that persons
attend to as explanations for this or that, where they get them from, which of
them they take to be exclusive of others, etc.

One of the ways of seeing how persons attend to explanations, etc., is to
consider when, for them, is a problem seen. Following is a quotation with a
similar cast to the foregoing. A teenage daughter in a family is having some
troubles; a social worker went to the family and spoke to them, and is now
reporting. At one point she says:

There’s no father. Her father’s been dead quite a few years and there’s

an unaccounted for, about three-year-old child in the home. . .

Now none of us take it that the social worker is suggesting the possibility of
a virgin birth. An unaccounted for three-year-old child is a child for whom
there is, not no father, but no husband.

What makes something ‘unaccounted for,’ and the kinds of things that
would account for it sufficiently, completely, and exclusively, are then matters
which we must very carefully try to get at. And I take it that there are very
strong ways that those things get regulated.

Now I want to continue in part with the discussion of the ‘warning,’
moving, however, a little bit further in the conversation. What I’m going to
say is extremely simple and obvious, but it will be of some use to see it and
be able to use it, because it has some general applications.

Ken : I still say though that if you take if you take uh a a big fancy car

out on the road and you’re hotroddin’ around, you’re- you’re
bound to get you’re bound to get caught, and you’re bound to
get shafted.

( ): ( , // >

Ken : W 1 look- I m gonna-

Al\ Not unless ya do it right. // That’s the challenge- that’s the
challenge you wanna try and do it right so you don’t get caught!

Roger : That’s the problem with society.

There are two returns to the warning, directly. One is, “That’s the problem
with society” and the other is, “That’s the challenge. You wanna try and do
it right so you don’t get caught.” Now one thing that seems relevant about
a warning - perhaps by virtue of the fact that warnings and predictions use

358

Part III

the same form - is whether the proposition it asserts is correct. And from the
two statements we have, there are some very simple and very nice ways of
seeing that Members accept the correctness of what might be a possibly
correct statement. One very obvious one is that they complain about what’s
asserted. The use of a complaint (e.g., “That’s the problem with society”) can
then be seen to be one way that a statement’s correctness is accepted.

I want now to see whether that assertion - that a possibly correct statement
is accepted as correct if it’s complained of - can be generalized. One way it
might be generalized is to consider some class of which ‘complaining’ is a
member, and to see whether, for some class of which it’s a member the other
members also do accepting of the correctness of a statement, if they’re used.
One obvious class of which complaining is a member is a ‘contrast class’ in
which another member is ‘applauding ’/‘praising. ’ 1 assert that; I’ll try to
show that it’s so because I don’t want to rely on the fact that the dictionary
says complaining and applauding are antonyms. Further, if all or some
members of that set do, in part, the job of showing an acceptance of the
correctness of a possibly correct statement, then we have a series of ways in
which such an acceptance can be located, and thus, that a statement is seen as
correct for Members can be warranted by us, i.e., we can point to the fact that
it was complained of or applauded. There are, then, two kinds of related
problems. One is to show that “That’s the problem with society” is a
‘complaint.’ and the other is to establish in some way the kind of intimacy
that we’re proposing holds between complaining and applauding.

In the case at hand, the problem of showing that something is a ‘complaint’
is not much of a task at all. The character of “That’s the problem with
society” is not uninteresting, but it’s clearly a ‘token complaint.’ If anything
is a ‘complaint’ this is one, and it’s simply recognizable as ‘a complaint.’ That
is, it’s a preformulated complaint. By ‘preformulated’ I mean that that
statement, “That’s the problem with society,” on an occasion in which
somebody wants to do complaining, can be used, and can be used about a lot
of things that they want to complain about. It isn’t restricted to a usage for
the particular thing that was complained of here. And we needn’t suppose
that it was constructed for this particular complaint; it was simply invoked.

That’s what we mean, in part, by ‘tokens.’ They are like money in that
sense. If you want to do a certain thing that can be done with this, you do it
with it. Now that matter has some very general import, as follows - and the
point that I’m going to make is Parson’s; at least he makes the observations,
and it’s a matter that he’s attempted to deal with in his inimitable fashion.
One of the nice questions that comparative sociologies obviously might want
to deal with is: For various types of currencies, what kind of distribution is
there across cultures? That is, if you can make a list of things that are doable
via tokens in some place, how many of them can you find in any given other
place, where those things will tell you a great deal about various kinds of
social development. For example, it’s obvious that if they have a token
money, then that’s a tremendously indicative fact about the kind of economy
they have. Now there’s no reason at all to restrict that notion to money. You

Lecture 12

359

could perhaps find out a great deal by virtue of whether they have, e.g., token
complaints, token praises, token etceteras. And also, of course, you’d want to
know the domains in which such things can be used in any given society.

Tokens are especially well-formulated parts of a place. They are routinely
used, and are designed for routine use. They don’t have to be constructed at
any given time, and the task of their acceptance does not turn on things like:
Do you have credit at this place, or Do they like the item you’ve built, or Can
they recognize that it’s the kind of thing that passes? It may be that anybody
can invoke them, and when they’re invoked, what they’re doing is perfectly
well observable.

Now then, to the question of the proposed intimacy between complaining
and applauding. It turns out that there is an intimacy, and it is extremely
nicely used. In certain ways it’s utterly crudely obvious, i.e., if you have a clear
member of one, then you can regularly make a member of the other by simply
negativing it: “He’s nice’’ (praise, applause), “He’s not nice” (complaint).
That’s altogether too obvious and simple and there are some less routine,
perhaps much more elegant usages, but that one will provide in part for the
alternation of complaining and applauding.

There’s a kind of thing which some people do with immense regularity,
which, if you ask them, they tend to say they’re doing perhaps to soften the
blow of a complaint, perhaps to show that they are not just complaining. But
for our purposes what it ought to be seen as, at least in part, is ‘signalling a
complaint. ’ What they do are things like this - and the following three
instances come from within about a minute and a half of conversation
between two people talking about a bridge game; produced by the same
speaker:

B: Yeah. She said then will you look at your hand. I mean, she’s very

nice, but to me that’s inviting you to gain.

Later on:

B\ I like her very much, but she still has that silly chatter about her.
And a bit later:

B\ Yeah, see we waste an awful lot, and I know- I think- Really, I love
Theresa. I didn’t mean that- her food was just outstanding. But I
thought we were going to eat at the tables.

Apparently, then, routinely a piece of praise plus ‘but’ plus something else,
tells you that the something else is a ‘complaint,’ where it isn’t obvious, often,
that it is a complaint. That is to say, isolated, it wouldn’t be a complaint. You
wouldn’t recognize it as such if all you had was “To me that’s inviting you
to gain,” or “I thought we were going to eat at the tables.” (In the latter case,
the problem was that this lady didn’t want to stop the bridge game to have

360

Part III

a meal, and her complaint involves that while the meal was marvelous, it
interfered with the course of the game.)

That thing is really extremely widespread. It seems to work the other way,
as well: Possible complaint plus ‘but’ plus something else, tends to be used,
I think, to signal ‘praise.’ “He really isn’t very cute but he took me to
dinner,” “On first meeting one might think of him as gruff or aloof.
However, in a short time his warmth and closeness to his family become
apparent.”

So I take it we can say that complaining and applauding stand in some
strong alternation to each other. And of course under that circumstance, the
sheer fact that we know that a ‘possible fact,’ if responded to by a piece of
praise is thereby accepted as correct, is not now a matter independent from
that thing about complaints, but is simply a rule formulatable by reference to
that class of which complaining, applauding, and the like, are members. They
are, among other things, ways of accepting the correctness of a possible fact,
and we sociologists can use their use to propose that some possible fact is
accepted as correct by Members.

And we can consider variations on the same sort of thing, which can turn
on such statements as: “Women are fickle,” return: “It isn’t their fault.”
Where, if your concern is to deny the fact, then “It isn’t their fault” fails that,
in part. Where, then, things like ‘excuses’ and other such things work at the
same sort of task, and we can then get ourselves a rather decent armory for
seeing the sort of thing that Members take to be correct, at least on any given
occasion some Member takes it to be correct.

Now the matter of the other fellow seeing the statement as proposing a
challenge, seems to serve, in part, the same sort of work, but in the case at
hand and for cases like it, what seems to be involved is that not all of the
possible forms that that statement might be heard as, are accepted as correct.
That is, the statement, “If you go hotrodding you’re bound to get caught”
can of course have a variety of interpretations. For example, just vary ‘you’
between singular-plural-everyone use. Given that “you’re bound to get
caught” could get various interpretations by Members, one of the things we
want to be able to see is, which ones are they seeing, and which ones are they
accepting.

When it’s seen as a ‘challenge,’ it seems that what’s involved is that it’s
seen via ‘you’ as a rather generalized term, and that “you’re bound to” is not
for any given case, but something like ‘over the long or short run,’
‘eventually.’ Otherwise, again, if it’s guaranteed that on any given try you will
get caught, then there’s no challenge

There’s a certain beauty to seeing a challenge in such a statement. And it
sets up some rather regular social problems of sorts. That is, if it’s very likely
that anyone who does it will get caught, then a game can be built under
certain circumstances. Or, even if you don’t fully build a ‘game,’ what you
can do is set up circumstances where you can see that you have succeeded, and
you can count each success. Where one has these sorts of statements accepted,
then, one has machinery which permits you to score, and to tell others, and

Lecture 12

361

they’ll see that you scored, i.e., you did it and you didn’t get caught. “I did
it five times and I didn’t get caught.” “We did it twice that night and we
didn’t get caught.”

A fairly full-fledged ‘game’ can be developed when that event that’s
troublesome (like getting caught) is the result of something that’s done by
some set of persons who operate under a set of rules that are laid down (the
police, for example). And it doesn’t matter if a set of rules are intendedly
regulating the challenger’s behavior, it can be read as regulating the behavior
of the catchers. If the challengers then treat those rules as enforceable on the
behavior of the catchers, then the catchers are bound by a set of rules that
permit them to be players in an indefinite series of games. The thing becomes
a game by virtue of somebody setting up another team which is going to treat
the rules that intendedly, perhaps, regulate themselves, as regulating the
others. They then set up a series of things they’re going to do, via which
they’ll score, which they can compare each other on, etc., and which can work
insofar as they can take it that, e.g., the police will abide by the rules of the
game. The police sometimes buy this; sometimes it infuriates them. They can
perfectly well know what’s going on. It’s the basis for complaints about
having been arrested and convicted, perhaps, for doing something which you
weren’t in fact located by virtue of having done. You were located only
because you were recognizable as an intended player in some game. They
didn’t let you play. They see you’re a hotrodder, and they know, for example,
that if they care to they can arrest you on whatever, and find something wrong
with your car, and that will then provide for an arrest having been done in the
first place. They do that, rather than awaiting that they catch you under the
ways that they’re ‘supposed to’ catch people. Where, for them, it isn’t the
circumstance that everybody is playing the game. If everyone were, matters
would be altogether different.

Now this building a game out of such sets of rules which are binding on
one set of parties (allegedly, perhaps, the ‘wrong’set of parties), which can be
publicly known, so that you have a rule book which doesn’t tell both sides
what each must do, but only tells one side what it must do, and therefore
permits a game to be built (or a whole set of them) is, of course, a
tremendously regular kind of thing, completely apart from situations with the
police. There are problems, of course: No referees. Which is the issue that gets
hotrodders into trouble, i.e., the police can violate and you can’t complain.
And of course the scoring is not exactly similar to the way scoring is done in
games, i.e., both teams do not agree at the end of an evening that you scored so
many, though the police may well on some occasion say, “You won this time.

And for a good game it’s of course necessary in the first instance that there
be another team and that they be doing things. No opposition, no game.
T. S. Eliot’s attack on modern sex in The Wasteland, center section, involves
the destruction of opposition. The woman doesn’t oppose, doesn’t do
anything, and then it’s a quite different kind of phenomenon. You can’t prove
yourself a man that way. You can’t score. Though you could tell your friends
you did.

362

Part III

So it isn’t, then, incidental where one is going to find systematic attempts
at violations of certain sorts of rules. There has to be some very good
enforcement procedure going on, where you’re going to get sport made. And
of course none of them see themselves as professional criminals; they are
sportsmen. And it just wouldn’t do to violate laws that nobody could see were
regularly enforced. That could be an argument for various sorts of things -
loose enforcement of the speeding laws, low speeding laws or maybe high
ones, at least if certain matters turned out to be relevant, like the death rate
among hotrodders is very high and you wanted to cut that down at all costs.

One further thing. On the matter of ‘complaints,’ there are several other
things involved when one uses a complaint to accept a fact. One of them is,
again, perfectly obvious and simple: The one who uses a complaint can be
heard as making an identification of himself. And that is obviously an
enormous control on making complaints. It may be extremely difficult to be
able to formulate a complaint about certain matters without having a
perfectly standardized identification asserted (e.g., ‘traitor’). Under that
circumstance, the counting of the amount of complaints versus the amount of
applauses has rather serious weaknesses. Clearly enough, you can’t do one,
perhaps, like complaining, without having a certain identification, or some
small group of identifications, placed on you in such a way that you can’t
readily get it off, and that itself may be relevant to the question of whether
you do it or not.

And on the other hand, the way you get such a thing put on you may be
either to complain or applaud. As a measure, then, of ‘response to the fact’
it has some weaknesses which are systematically built right into the very
mechanism; that is, those other things that can be done with complaints or
applause.

Lecture 13

Button-Button Who's Got the Button

I felt a strong need to take a break from the data which we’ve been dealing
with. What I’ll consider today will be the first in a series of interjected lectures,
which turn out to be quite relevant to certain matters that are present in these
materials. For example, I very roughly suggested the character of hearing the
‘warning’ as a ‘challenge,’ as being involved in seeing the possibility of a sport,
or a game, in the situation of driving under certain regulations. What I want
to do, then, in the series of lectures I will give occasionally, is to develop a
formulation of series of the properties of games. And I’ll try to point out now
and again what aspects of the data we are dealing with, that they are
important for.

I’m going, now, to consider some aspects of a particular game. By no
means anything like all aspects of games, or all relevant aspects of that game.
The game is called Button-button who’s got the button - an enormously
substantial phenomenon, you might imagine.

I’ll start ofF by just going through how the game works, roughly. It’s
played by kids five to ten, and it works in the following kind of way. There’s
a set of people sitting in a row on a sidewalk or something like that. There’s
one person who is called It, and they walk down the row with a button
concealed in their hand, and they engage in possibly passing it to each person
in the row. Once a cycle is completed, the person who is It having possibly
passed it to each person, then various people in one or another way, guess who
has it. When the guess is correct, then that play ends and they begin again.
There are alternatives that can be done; perhaps the person who guessed
correctly is It, alternatively, the person who had the button can be It. They
then proceed to do the same thing, and the game goes on in that way.

Now, I take it the game is tremendously interesting for certain sorts of
tasks that are involved in it. That those tasks are learned to be accomplished
via the game alone, I’m certainly not saying. That the game can be a large
help, and that other games and activities like it can be a large help to learning
those tasks, I think is so. That those tasks are altogether fundamental to
becoming a full-fledged human seems, I think, to be clear.

I’m going to try to provide a characterization of what’s involved in playing
this game, or playing it well.

The problem for a person who is engaged in trying to decide who has the
button - where that person doesn’t have the button himself - can be said to
be detecting, in what may be a series of alternative ways, from basically the
appearance of the other non-It players, which of them has the button. The
alternatives are, for example, very carefully watching the moment of possible

363

364

Part III

transaction and/or looking at them afterwards and seeing if there’s some way
to see that they do indeed or do not indeed have the button. Now the physical
properties of the game are quite relevant, in that a button is not an incidental
phenomenon. It is very small and thin, and can then be passed in such a way
that when you see that someone has the button, you don’t see the button. So
we get a notion of observability which involves something like ‘seeing that
they know they have the button.’ And the game, then, is rather different than
a game in which, for example, a piece of food might be passed and one could
smell who has it, or the object that was being looked for had some visible
effect, like the hand was thicker than it might be if it didn’t have the button
in it; where, then the button itself would be the observable.

But what’s core to this game is that one can observe that someone has the
button, feel very confident in that, without having any observation of the
button itself, and it’s simply a failure of the passer or the recipient that it’s the
button that’s seen and not their appearance which gives away that they have
the button. So it’s not the case, then, that the object is incidental to the
characteristics of the game and that it just happens it’s played with a button
- though of course it is in some sense just incidental that it’s played with a
button; it could be played with anything that had the same physical
properties. But it couldn’t be played with anything that had different physical
properties. Something that’s thicker would not do the job and would pose a
very different kind of game.

Now this sense of an observable thing - that one sees an appearance and
sees, then, that something has been done - is of course in its fashion, the basic
sense of ‘observable’ for social phenomena. Especially for, let’s say, moral
phenomena. One can see that somebody did something wrong by looking at
them. One can see who they are by looking at them, etc. Where it’s not
anything like the smell or the size of the particular object. And in that sense,
the notion of ‘observable’ is that one sees their thoughts. That is to say, one
sees the button by seeing that they know they have the button. If the button
could be passed in such a way that they didn’t know they had it, the game
would be altogether different.

And this phenomenon of seeing other people’s thoughts is really an
important thing. Exactly how it’s properly posed is quite tricky. First of all,
it’s of course nonsense to say that thoughts are things that can’t be seen, unless
you want to take some notion of ‘thoughts’ which Members do not employ,
since they certainly do take it that one can see what anybody is thinking. Not
in every case, certainly, but you can see what people are thinking, and there
are ways of doing it. And you must learn to do it.

The question of whether you learn that others can see what you’re thinking
and you learn to see what others are thinking, or that you know it in the first
instance, is what’s problematic. I’ll do a parenthetical series of considerations
on that now.

One of the classical symptoms in certain kinds of schizophrenia is that the
patient takes it that other people can see their thoughts, or know their
thoughts - look right into their head. Adults who hear that take that as an

Lecture 13

365

immensely queer symptom. And the problem, at least historically, was posed
as: How is it that somebody would come to feel that others would know their
thoughts? That’s a very obvious way of posing the problem. Somebody says,
“Others know my thoughts.” How do they come to think that?

One of the early places, in at least the psychoanalytic literature, that that
formulation was used, was in a very classic paper by a man named Victor
Tausk, called ‘On the influencing machine in schizophrenia.’ It’s been
translated and reprinted several places, including The Psychoanalytic Reader
by Robert Fliess and Psychoanalytic Quarterly. It was delivered in about 1917
at one of the meetings of the Vienna Psychoanalytic Society, and at the end
of it there’s a discussion. Tausk reports that he has a patient, a 1 6-year-old
girl, and she has a very characteristic phenomenon which he discovers in her.
He’s talking to her, and she laughs at his questions. He asks her why is she
laughing; she says, “What are you asking me questions like that for? You
know what my thoughts are.” And now he tries to formulate a solution to the
problem, posed, as I’ve said, as ‘how could somebody think that others know
their thoughts?’

Freud is at the meeting where Tausk delivered this paper, and with what
could be seen as a very characteristic aspect of his genius he says, “You posed
the problem in completely the wrong way. It’s not the case that what we have
to explain is how is it that people come to think others know their thoughts.
That’s not the issue at all. The question is, how is it that normals come to
think that others don’t know their thoughts? And the character of the
schizophrenic as ‘regressed’ involves the fact that either they never learned
that others don’t know their thoughts, or, having learned it, they are now
back at positions akin to that of a child.”

And he formulates the following kind of account (I’m not selling this
account, I’m only offering it to loosen the matter up somewhat). He says,
“After all, children learn their language from adults, and they must suppose
that as they use it, adults know how they’re using it, what they’re doing with
it, and can directly see their doing that sort of thing, i.e., thinking with a
language,” Now he says, “Well, what will be the sort of thing that would
lead to the discovery that people don’t generally or in all cases, know what one
is thinking?” Then he says, “The first successful lie. What a trauma that must
be. Or what a discovery that must be. You’re asked a question, you give a lie,
and they don’t tell you you’re wrong.”

That’s a tremendously interesting set of possible observations. And it
makes very equivocal exactly how we’re to formulate the state of knowing
that others know your thoughts, or knowing others’ thoughts. But in any
event, it’s clear that to some extent people take it that others can know their
thoughts, and people take it that they can know others’ thoughts, by looking
at them.

Now, if we consider the game of Button-button who’s got the button, we
could imagine that very young kids who are playing it, and who take it that
what they have to see, to make a decision, is the button, might discover that
they don’t have to use the button. That is, they could be spending their time

366

Part 111

watching for the button, fail to see the button, and nonetheless see that, gee,
they bet they know who has it - by whatever it is that seems to them to give
that away. So it’s the kind of game that may very well involve something
that’s in general terribly important, and that is that you can regularly - and
children can - start off with quite lousy criteria for making a decision, be
correct, and improve by modifying those criteria as you go along.

A very young child, for example, can have very good ideas on who’s boys
and who’s girls, and have very bad ideas on what differentiates those two.
Though in any given case they may make quite a good formulation of who
it is - a boy or a girl. And that then involves us in a situation where there’s
a strong separation between the adequacy of one’s criteria, and practical
efficacy. That’s of course, not simply at the basis of the possibility of
improvement to some extent, but of, let’s say, infantile socialization. And
presumably a great many things like science tremendously depend on it. One
can make differentiations which turn out to be wrong but have permitted you
to locate a phenomenon that remains a phenomenon as you get better at
deciding what it is or how it works or what differentiates it from something
else.

Now in playing the game, one has to come to see that others can tell from
your appearance that you have the button, and now you can make a
tremendously important discovery, probably akin to what Freud was talking
about as the first lie. And that is: There is something you can do which has
an effect on whether they guess you have it when you have it, or guess you
have it when you don’t; and that if you do those things you get a variation
in your name being picked or not. Where what that turns on is quite nice in
the game. The core thing is the point of possible transition. One might
suppose that if the kid were being altogether naive, then if he gets the button
he’ll respond dramatically one way or another, e.g., with joy or sorrow, at that
split moment when he gets it. What he has to learn is, for one, to be prepared
to control a response, depending on what happens - where he doesn’t know
in the first instance what’s going to happen. That is, if he wants to be
imagined to have the button when he hasn’t been passed it, he has to be
prepared to act as if he’s got it at the moment of possible transfer. But if he’s
prepared that response too well, then if he does get the button he’ll have given
away that he does have it. That means he has to develop a control which is
ready to work depending on alternative events - a control which works almost
immediately.

The usability of appearance, then, to convey a false impression, is one of the
basic tasks in playing the game well. It is, then, in part, a training ground for
liars and deceivers. But it is equally well a game which is directed to the
detection of liars and deceivers, since the category of player you are is regularly
changed and therefore you not only have to learn how to deceive, but how to
detect deception.

Furthermore, you have to learn that that which successfully operated once
or n times is not guaranteed success, and in fact may well be assured failure,
a next time. That is to say, if you joyously respond to the fact that you don’t

Lecture 13

367

have the button, and people take it that you have got the button and discover
that they’re wrong, then any next time they can learn your pattern and use it
to see what you’re doing. So there’s no winning strategy that can be laid out
on the basis of any past strategy. And then it’s the character of the game as
replayed constantly each time, which is rather crucial to training for both
deceiving and detecting deceiving. Any successful method for doing either can
be self-defeating if it’s simply continued.

Now one of the serious things you have to notice about the game, and this
holds to a very considerable extent for games in general, is the following:
Recall, in the piece of data we’re considering, A1 says “Now did you do it
right that’s the challenge you wanna try and do it right so you don’t get
caught.” One of the things we might be engaged in trying to figure out is,
well, would it really be a proof that you did it right if you don’t get caught,
or, for that matter, would it be a proof that you did it wrong if you do get
caught. If we took, say, science as our model, then it would be to some extent
irrelevant, and certainly not definitive, that you did or did not get caught -
for example, that you did or did not make a correct guess. What would
differentiate a guess from a prediction would be that you have, and can offer,
a good account of how you came to see that the person had the button, i.e.,
for one, how you came to do it right. Whereas in games, and in games like
this one especially, it’s asserting the name or getting caught that counts. And
one can have the most elaborate or the least elaborate systems of accounts and
they’re irrelevant. Games operate under a test of practical efficacy.

And that’s not a matter of theoretical preference for the players. If you lost,
if people have called your name out when you’ve had the button, you have
to suppose that they have some way of knowing it, and you have to modify
your activities.

Now you can apparently also proceed play by play, have a notion of success
that will serve you, figure that you’re learning and improving, without
operating in any situation where you could say “What I’ve learned now is
sufficient for never losing.” That is, at no point need you propose to test the
adequacy of what you’ve gotten by saying “What I’m going to do is simply
maintain the way I’ve got now of knowing how others look when they have
the button, or of having others not know when I have the button, and I’ll be
right.” In the game, that is one way of guaranteeing that you’re going to be
wrong, since it simply poses the problem for the others to see what pattern
you’re using and permits them to modify their own activities so as to foil it.

The fact that there is a shift of categories in each play, may be of some large
relevance as well. Many games have that property: We bat this half inning,
you bat next half inning, etc. And that involves, for one, that in simple games,
each position is learned from both sides, so that a very important notion of
categories and their characteristics is taught to kids then - if they don’t know
them beforehand. That notion is: Categories aren’t persons. There’s no
equivalence either of one category - one proper person, or that a person has
only one category indefinitely. This is a matter which they might well at some
point in their lives not be apprised of and which might, then, involve them

368

Part III

in formulating, for example, activities as tied not to categories but to persons,
which would be evidenced by their use of some activity characterized in a way
that, for adults, would not be properly formulated. That is to say, where the
appropriate category is not used.

You find, for example, in the younger children’s stories in Children Tell
Stories , that they seem to have basically just one membership categorization
device - the ‘family’ - which they use to handle almost anybody in any
situation there may be. Now that fact doesn’t just happen, because adults
allow them to do that. That is, children are trained first to use a single device,
and the device is modified from its normal uses so as to permit its use to
handle pretty much anybody. That is to say, the ‘family’ device has certain
possibilities which permit its application to almost anybody. Certain catego-
ries can only have a single member in them - ‘mother’ and ‘father’ - but there
are other categories - ‘uncle’ and ‘aunt’ - which in principle can have an
indefinite number of members, and are taught to children as having that
characteristic, so all adult males, or many of them, for some period of time,
are Uncle so-and-so without respect to any formulation of kinship as adults
would use it, and females are Aunt so-and-so. And in that way, the use of the
single device is permitted and encouraged, and it’s made into something that
can handle almost anybody.

And what we say if that’s so is, of course, that they’re using the categories
of the ‘family’ device as equivalent to personal names, where ‘personal names’
are in some way equivalent to, for objects, the name of objects. It’s only in
special circumstances that a table need be called anything else than a table.
Whereas for formulating who’s doing something, a set of alternative
categories may well be extremely relevant, and one presumably has to come
to see that a person can have a whole set of alternative ones. One very simple
way is to have a rapid shift in those categories that hold for someone, where
it wouldn’t be odd at all that someone was It a minute ago and now is no
longer It. That can be, now, a prototype of how it is that persons ought to be
formulated, rather than that Mommy was Mommy yesterday and now she is
no longer - where Mommy becomes stable by virtue of becoming a term of
address in a sense, and thereby can retain its character as being used
indefinitely perhaps, while one learns that there can be a series of other
formulations of the same person, for others and himself, depending on what’s
going on. We get, then, the various sorts of category instability, where the
stability resides in an appropriate category for an appropriate activity and not
an appropriate category given some person having had it.

I have some indication that there is in fact a discovery procedure of learning
a series of these matters, where they are, in fact, learned over time. For
example, it appears that in games like Hide and seek, when young kids are
playing, if they win once by hiding somewhere, then the next time they hide
in the same place, and they have to be taught by the other players that that’s
not a guarantee of success, it’s a guarantee of failure.

We could take these features and show that there are games in which just
what works here, the opposite works there. So, for example, if in Button-

Lecture 13

369

button what’s important is to control one’s future immediate response to
possibly getting the button, then there are games where, if a person does that,
that’s a good guarantee that they’re cheating. What they ought to do is to
respond in absolute naivete so that, for example, in the game Pickup sticks,
if the sticks move then your turn is over. And it’s not supposed to be the case
that the other person has to decide that the sticks moved.

The game is typically played where the other person may be looking, not
looking, paying close attention or not, and what they see is that you’re
watching the sticks very carefully, which is what you’re supposed to be doing,
and that you obviously react to the movement of the sticks. If they move,
your face shows it. And you just drop the sticks; you don’t have to wait to be
called on it. In fact, in that sort of a game, if you’re playing with somebody
who you suspect is a cheater, all you have to do is watch the sticks, see if they
move, and see if there’s any response on the part of the person whose turn it
is. If they don’t respond, then you can be pretty sure that they are engaged in,
systematically or not, cheating.

A few comments: This issue of the relevance of an appearance for attending
what’s up with somebody, and their capacities to control it and the equivalent
capacity to check it out, is one that I can only very generally, very abstractly
handle, because how it’s ever done is something that first of all I’m not in a
position to say. I don’t know that it would ever be an issue for Sociology to
say, as compared to that being a fact that students of Perception would deal
with.

I make the point now, and it can be generalized to a lot of what I’ve
already said, that once we get to an issue of a single device being relevant, then
we have nothing to say. I have no idea how it is, and I make no claims to have
any idea how it is. Once it’s the case that Negro-white is relevant for some
person, male-female is relevant for some person, somebody goes about
detecting that that person is ( ) I’m not ever going to be

saying anything about that matter.

On the orientation to appearances more generally; when I said that it’s
utterly central to being human, I had in mind the following sort of thing. In
our Judeo-Christian mythology, the first event characterized as being human
is a response to an awareness of an appearance. That is, the first event after the
eating of the apple in Eden is just that event of being aware. The first thing
they have- happen is being aware that they’re naked. Before that they’re not
human, and before that they’re not oriented to what can be seen by looking
at them. So the matter has been tremendously focussed on, at least in such
kinds of places.

Lecture 14

Disorder ability; Tying rules

What I want to do now is introduce another order of consideration which is
relevant to the data we’re looking at, and then, directly, to conversation in
general, and then indirectly to whatever activities one might want to consider,
apart from conversation.

I have now and again made some remarks about sequencing, and one
obviously relevant matter about activities, particularly conversation, can at
least initially be posed as follows. First, to what extent is it the case that for
any given conversation the utterances which compose it are not disorderablel
And by that I mean only that, whatever it is that one might want to preserve
or need to preserve so as to analyze what’s going on, can one, for example,
take any utterance, wherever it occurs, and place it somewhere else in the
sequence for the same conversation? “We were in an automobile discussion”
“discussing the psychological motives for” “drag racing on the streets.” Does
it matter that it goes “We were in an automobile discussion” first, “discussing
the psychological motives for” second, and “drag racing on the streets” third?
Or could you put it any way you want it for purposes of analysis, and perhaps
for purposes of having produced a conversation. Now it’s clear enough that
there are some parts of a conversation for which it might be the case that it
wouldn’t matter. Suppose a new topic is introduced in some place in the
conversation, and that topic goes on for some way and then closes, and then
another begins and closes. It’s possible, perhaps, that you could take the
second topic discussion and place it before the first, and find that it makes no
difference whatsoever. Or it’s possible, perhaps, that you could take any
utterance out of a conversation and place it any other place, or some other
places, while losing nothing. That would be one type of activity. They happen
to occur in sequence, but there’s no local control over when anything happens.

Second, if .it’s the case that there’s at least some partial non-disorderability
to the utterances of a conversation, how is it that that non-disorderability is
provided for? There are, at least a priori, quite different possibilities.
Supposing some non-disorderability for any given conversation, it might be
the case that the way that that non-disorderability is provided for is particular
to the given conversation. Alternatively, there might be a more or less general
set of rules which can be used to provide the non-disorderability of any
conversation at all.

For some purposes, obviously, disorderability matters. When we talk
about the fact that some statements in a conversation have to be understood
‘in context,’ one of the things we’re pointing to, presumably, is that to treat
them without respect to the problem of non-disorderability is to provide

370

Lecture 14

371

perhaps for misunderstanding, or to make them altogether not understand-
able. That’s a very familiar observation. Here’s a quotation, a sentence
President Kennedy once said:

Children should bury their fathers.

I suppose one could say, gee, that’s a strange thing that fellow said. If one gets
what he was doing, what it is that preceded that, one might feel that one was
in a position to find out what activity he’s doing with that sentence. In this
case he happens to be dedicating a memorial to persons who were lost at sea
during the Second World War, and he’s trying to indicate the pain in such
losses. The line I’ve quoted occurs in the following utterance:

It is, after all, against the law of nature for parents to bury their children.
Children should bury their fathers. And when it’s necessary for a father
or mother to bury a son who may range from 18 to 28 , with all of his
life before him, it represents a special wrench.

The line itself, however, may not have any formal properties which tell you
right off that it’s being ‘pulled out of some context.’ There are others which
tell you right off that they’re pulled out of some context to which they have
to be returned to find what they’re about. What sort of line? How about
‘Yes.’ ‘Yes’ what? Look to the question above, perhaps, to find.

Now of course there are various consequences to treating properly
non-disorderable statements as disorderable. One is, if statements are alto-
gether disorderable, then they are apparently excerptable and quotable. And
the issue of quotability has been a rather prominent one in many places. One
can think simply of such matters as indicting somebody for some statement
they made. But there are much larger considerations involved, and, in part,
the difference between Aristotle’s techniques of doing philosophy and Plato’s
turn on the fact that the character of a dialogue as a method of doing
philosophy has as one of its central properties that the utterances of the
dialogue are not properly to be pulled out and merely, then, treated as what
this fellow thinks about a problem that this sentence happens to be about. In
part it is that sort of fact that makes for the difficulty of analysis of dialogue,
the sort of inaccessability it has, the possibilities of developing a position, and
various other things.

Now if we look to conversations to see to what extent any one of them is
disorderable, or not, and then ask, how is it that that disorderability or
non-disorderability is provided for, we can perhaps come up with some rather
general sets of rules whose use does the work of providing for non-
disorderability. And if it’s the case that we need some such rules and that if
we had them they’d give us the non-disorderability, then, for one, we’d have
a way of formulating some decent notions of ‘context,’ seeing that context is
something that is oriented to by Members, and is not then simply a matter of,
‘Of course it happened to happen at a certain time, so therefore you can’t

372

Part 111

remove it from when it occurred . . etc. In many areas there are significant
attempts made to produce utterances which are not context-bound at all for
their understanding. Obviously to some extent the whole science of mathe-
matics is directed to that sort of attempt. And of course the very notion of
‘truth’ has such an intention involved in it, though there is a considerable
amount of argument as to whether such a claim can be made.

Let’s then just consider a little bit, a few of the ways that non-
disorderability can be provided for, and then I’ll offer some rather general
considerations about it. I’m going basically to be talking about two classes of
rules which provide for non-disorderability. I’ll use the term I introduced
earlier as a general name for these rules, 'tying rules.' And the two classes I’ll
call first speaker rules and second speaker rules. Both of them concern pairs
of utterances. The basic difference is, for the analyst anyway, that for first
speaker rules, one sees, when looking at some utterance, that its user has made
himself into a ‘first speaker’ for some pair, and has provided, then, that a next
speaker ought to make himself a ‘second speaker’ for that pair. One starts
with a recognizable ‘first member of a pair,’ where it may well be that if you
don’t know that there has been a first member of a pair you can’t tell that the
second is ‘a second member of a pair. ’The second members are not
recognizable, the first are. The second are only recognizable given the fact you
recognize the first and are therefore looking for a second. That is, the rule
doesn’t govern the first speaker’s actions.

For second speaker rules it works the other way. You start with something
recognizable as ‘the second member of a pair,’ and then turn backwards to
something that is not recognizable as ‘a first member of a pair’ unless one has
the second to see that there ought to be one. So first speaker rules provide for
the proper tying of a next utterance to a current one, and second speaker rules
tie some second utterance to something which may not have been produced
as an intended ‘first utterance’ for some pair.

Now the use of these two classes of rules is not mutually exclusive. That
is to say, a speaker can do both at once. He can make his utterance a ‘first’ for
some next, while at the same time making it a ‘second’ utterance of a prior
pair. And that’s extremely important for the following reason, for example:
One of the most obvious first speaker pairs is question-answer, the ‘first’ item
being a ‘question.’ Now the first item, the question, can also be the ‘second
utterance of a pair’ in the following sort of way. Suppose one person has said
“My opinion is X.’’ Then we could have a question which would say
something like “Why do you say that?,” where that’s partially tied to the
prior, and provides for another.

The theoretical core interest in those things is that if you can combine the
two types, you can get right off, triplets, and an indefinite nesting of a
conversation which ties all of its parts together in some extremely complex
way, out of very simple pairs of rules. That is, you might well find that some
item is the second for some first, and that that first is now tied to some other,
and back, so that if you want to put a given one anywhere, you find that you
have to put it exactly where it is — and not just following the immediately

Lecture 14

373

prior utterance; it would have to come after the eighteenth utterance in that
particular sequence. And all you need for that is, perhaps, two pairs of rules
plus their combinability.

Now what are the sorts of rules? One obvious first speaker pair is, of course,
question-answer. By using a question, a speaker makes himself a ‘first
speaker’ and provides that somebody ought to be a ‘second speaker’ for that
pair. It’s routinely the case that questions are produced in such a fashion as to
be recognizable as questions. There are a variety of means whereby that is
done, which are, of course, not particular to any question. They are things like
question-order in sentences - a different grammatical order than declarative
sentence order, intonation pattern, and a series of others. There’s a rather
lovely book on questions called Interrogative Structures of American English,
by Dwight Bolinger. It’s publication no. 28 of the American Dialect Society
(1957). And of course there are routine grammatical discussions of them in
almost any grammar.

Let me pause at this point just to make a few remarks on questions; I don’t
intend this as a comprehensive discussion. It’s occasionally or regularly argued
that certain criterial features of questions, for example, in intonation, the
characteristic rise, are present far less frequently than is regularly supposed.
For a discussion on that matter there’s an article by Charles Fries in a book
called Essays in Honor of Daniel Jones. What Fries did was to listen to the
television program ‘What’s My Line,’ take all those things which turned out
to be treated as questions, and try to write down their intonations, for one. He
came up with a finding that only something like only 3 5 percent of them had
the characteristic intonation structure of questions.

Now that doesn’t necessarily mean that questions are problematic. One
thing it might mean — and this is an extremely rough hypothesis, but my
suspicion is that something like it is so - is that what’s troublesome about that
sort of analysis is that essential to the argument is a notion of the
independence of any question from any other question. It seems to me from
the tapes that I listen to, that the chaining rule for questions is terribly criterial
to the recognition of questions. Such that the first question in a sequence will
regularly be independently recognizable as a question by, for example, a very
pure grammatical form and/or a characteristic intonation, but that once a
sequence of questions is started, then it may well be the case that unless some
next utterance by, say, the questioner, has a very clear non-question form, it
can be heard as a question by virtue of occurring in a sequence of questions.
Under that circumstance, i.e., that the characteristic of a question is to be
found by its occurrence in a list that is hearable as being ‘a list of questions,’
it would be difficult to warrant an argument which counted them as singly
independent objects.

Questions are, then, by and large, recognizable as such. And recognizable,
then, as first items in a pair. Answers are not. There are extremely few answers
which, if you simply have them, you know that you have ‘an answer.’ I once
tried to see whether there were any utterances that could be said, simply by
looking at it: ‘If that occurs it may always — or overwhelmingly if not

374

Part III

always - be an answer.’ And I came up with a kind of a funny result. I don’t
know whether it’s really correct; if it’s partially correct it has some interest,
because the item that I tended to come up with was one that not only may
well be always an answer, but it may also be an item which is almost always
offerable as an answer, i.e., can be offered pretty much to any question that’s
asked. And that is the statement, “I don’t know.”

What’s interesting about that is that it may permit us to partially explicate
a rather fundamental historical fact in Western civilization. And that is, that
“I don’t know” as an answer, has played a unique role in the history of
Western civilization. It was, of course, Socrates’ phrase. And it served for him
as a warrant for fundamental reconsideration of knowledge. It also was the
programmatic answer he sought to achieve in any given dialogue, i.e., to get
his respondent to offer.

Now, if a dialogue composed of questions and answers is going to be used
in an attempt to reopen what it is that is to be allowed as ‘knowledge,’ it may
not be, then, altogether incidental that the item that was used was one which
can be offered for any possible piece of knowledge and is, then, general to any
issues of knowledge. No other starting point might have that kind of
universality to it. I’ve asked people whether in ancient Greek it’s the case that
“I don’t know” is usable for any question - because it could perfectly well be
the case that it’s doable in English but not in Greek - and they said Yes. And
if that’s so, then we could come upon a rather nice basis for, in part, Socrates’
use of conversation, and particularly of the dialogue, that wouldn’t be
independent of this item, “I don’t know.”

Let me turn to some second speaker rules. I suppose in doing these we
could be getting into some kind of trouble with grammarians, but it’s a
trouble that with mutual decency can be avoided. But they by and large tend
to restrict their remarks and analyses to the single sentence as a unit - or the
clause, which could be equivalent to the sentence. And while they might
propose that the rules they use are similarly restricted - and we’ll propose that
the rules for use in grammars are not so restricted - we could easily say that
they might not be entitled to say that the rules are so restricted, since sentence
sequences are rarely studied by grammarians, and we won’t bother much
about the use of the rules in single sentences.

Some of the second speaker rules are very conventional rules of English
grammar - of a sort, anyway; though even they have been sometimes queerly
treated by grammarians. For example, there’s a rule like N ■*— P. ‘N’ is noun,
‘P’ is pronoun. The arrow is going backwards because, while ‘N’ occurs first
in the sequence, we start with ‘P’ as the thing that tells us we’ve got a case of
the rule, and we then look back to find the ‘N’ for that P.’

Now it seems to me perfectly obvious that that rule [N - *— P} is not
restricted to a use in a single sentence. It can occur in a single sentence, clearly:
John put on his hat. ‘His’ and ‘John.’ But it can be used across utterances, and
across conversations. I’ll show a very simple use of it across conversations, to
give some idea of its importance across conversations. And with it we get one
of the ways that we can even begin to dream of talking about ‘non-

Lecture 14 375

disorderability’ for a history of conversations between some persons; a matter
we would perhaps much like to achieve.

It’s kind of a long quotation. It starts out with a mother talking to a little
girl.

The mother asked as if wanting to get things straight: “What are you
going to say to Otto?”

Maude said patiently, as if she knew what were expected: “I’m going
to ask him if he can come over.

The mother added firmly: “And you’re going to say that you can’t
come over there if he asks you.”

Maude nodded acquiescence pleasantly and said: “Ok.” She turned
and ran quickly into the living room. She climbed onto a chair, stood on
the seat, and leaned her tummy against the back of the easy chair. She
reached up and took the phone from the top of the piano. She set the
phone on top of the chair. Maude waited a moment staring absently at
the chair back, then she absently said: “8—5.” She looked over at me
and asked curiously: “What are you writing about?”

I said to her: “I’m writing now that you took the phone from the
piano and said 8—5.”

She smiled and looked embarrassed. Suddenly she said: “Otto?” in a
very demanding tone of voice. Now her attention seemed to be quite
involved in the phone conversation.

It goes on and on. Then we finally get:

Mrs Pinter called, with a little concern, from the other room: “What did
he say?”

Maude called back impatiently: “He hasn’t telled me yet.”

We clearly have here two different conversations between the mother and the
child. For one, there are two intervening conversations, one with the observer
and one with Otto. (And one could in fact find the same sort of thing where
the two conversations take place on different days, for example.) And in this
case, we clearly have a use of N •*— P (where ‘Otto’ is ‘he’) which is occurring
between two conversations.

In the 1 1-page group therapy segment, many uses of this rule can be found
which are not within a single utterance. Very largely they tend to be in
juxtaposed utterances, but there are some characteristics which provide that
they needn’t be in juxtaposed utterances. For one, this particular rule has an
indefinite extendability to it. That is, given N ■*— P, you can just continue with
P, each P referring back up the sequence, and you can get long strings tied
together from the first N; that is: N P P *— P P etc.

Lecture 15

Tying rules; Playing dumb ;
Correction— invitation device

I’ll continue talking about these tying rules. My main interest is not in great
long lists of them, because they’re kind of dull things to make lists of, but in
what they do. I’ll consider a few; one can make, I suppose, hundreds more if
one likes.

I’ve been talking about the second speaker rules, and 1 talked of them via
the analyst’s tasks, where for second speaker rules one begins with what is
sequentially the second item but is recognizably the first that you can find.
That’s, then, the N •*— P rule. Occasionally you can see in some noun usage
that what you’re going to get is a case of the noun-pronoun rule the other
way, i.e., N P. That would typically be within one single utterance. For an
example of that, consider the line, “In that Bonneville of mine. I could take
that thing out. . .” When you hear “In that Bonneville of mine” at the
beginning of an utterance, you can be rather assured that in due course you’re
going to get a pronoun for “Bonneville.” Now that’s grammatically obvious
in that what one has is a partial disordering of the standard form of the
sentence; the utterance begins with a prepositional phrase where ordinarily,
let’s say, a sentence might close with it, and given that it begins with it, then
when the object that’s being referred to will be re-referred to, it will be
re-referred to via a pronoun of some sort.

Now there’s a function for that sort of thing, which is, for one, that
reversals of some of these second speaker rules can be used to preserve the
openness of an utterance, i.e., to indicate that at some point where it might
be closed, it isn’t going to be closed yet. And there are a variety of such uses.
Another, for example, involves, not the use early in an utterance of a noun
which looks surely like it’s going to be replaced at some point by a pronoun
so that you at least wait until that happens to find that the utterance is
complete, but the early use of a pronoun - where it doesn’t appear that it’s a
case of N •*— P, i.e., a pronoun tied backwards. For example, in this
statement, “Ken, face it. You’re a poor little rich kid,” where you have a
reversal, starting with the pronoun, and in that way this statement is seen as
not complete until what follows occurs.

The problem of preserving utterances as open is a systematically important
one, given the fact, for use, that there are a variety of things that can
stand - simply by, e.g., grammatical features - as complete utterances.
Sentences can be complete utterances, and if one wants to build one and
indicate that it’s not a complete utterance, then there are ways of indicating,

376

Lecture 15

377

in perhaps your use of grammar, that you’re keeping the utterance open.

Now apart from the noun-pronoun rules of the second speaker form,
there’s a corresponding one for pro-verbs, which is verb pro-verb. For an
instance of it, look at that statement “if you’re hotrodding around you’re
bound to get caught,” followed by “Not unless you do it right. . .you wanna
try and do it right so you don’t get caught!” Where ‘do’ is the pro- verb for
‘hotrodding around,’ tying the second utterance to the first.

On that use you have to be fairly careful, since ‘do’ is, besides being a
pro- verb, an ‘auxiliary,’ and one can have a verb phrase with an auxiliary in
it like ‘do,’ and have a second utterance which involves just the use of the
auxiliary: “Did you bake a cake?” “I did,” where the first ‘do’ was an
auxiliary, and the second is now part of that auxiliary used to do that whole
job of re-referring. There are lots of such things, in which a whole verb phrase
which includes an auxiliary has a second use in which just the auxiliary is used,
and that’s another class of tying rules; for example, there’s one which goes,
“Eventually you’ll become sane.” Then, “Or your mother will,” which is a
next utterance tied back to ‘will become sane’ as the verb phrase. The use of
the auxiliary to tie, and not the whole verb phrase, is, again, a very regular
kind of thing.

To see that there are an enormous number of such rules, recall that at the
beginning of our discussion of this data, I introduced the ‘completer’
situation, which was my first second speaker rule, i.e., a second speaker
beginning his statement in such a way as to make his statement a dependent
clause to what he makes an independent clause, but what might have been a
sentence; for example, that business of ‘discussion-discussing’. And if you
take all the ways that dependent clauses can be built to make sentences
independent clauses, you have a tremendous list of second speaker tying rules.
For example, you can build something that stands as a prepositional phrase to
a possible sentence. There’s one that goes, “It’s a good place to spend
Saturday morning,” where simultaneously, then, two prepositional phrases
are built: “If you’re stupid,” and “after Friday night” (data handout, just
below 9:30). So there are an enormous number of these type of second
speaker rules which are formed of rules which have their supposedly typical
place within a single sentence or single utterance. They need not at all be so
restricted, and you can find them all over the place. There are also, of course,
many first speaker rules: question— answer, request-return (acceptance or
rejection, whichever it may be), command-return, insult-return, etc. If you
go through a conversation, either building rules or just using rules you know,
and then try to see how much of it is untied, it’s a decent kind of exercise, and
in the particular conversation we have, you’ll find extraordinarily little that is
not a tied utterance in one way or another.

Now there are rather different ways of deciding, perhaps, that you have
something which is provided for by each set of rules. One way you might go
about deciding that you have a first speaker rule item is kind of useful in that
it suggests one consequence of first speaker rules which might not be
altogether obvious. And that is, if you’re dealing with a multi-party

378

Part III

conversation, then the use of a first speaker rule not only regulates properly
the speech of the person who ought to follow it with some second part to that
rule, but it also regulates properly, those who are not so told to speak. And
one can see that in operation via the ‘interruption’ notion I offered earlier, 1
which is that if what we take to be a first speaker rule is used, addressed to
some second speaker, then if somebody else starts talking before that second
speaker does, and the second speaker begins, then one regularly finds that the
person who started talking first stops, and allows the second speaker to
proceed in conformity with the first speaker rule. Here are a couple of
instances. In the first, A, talking to B, says “Hey don’t tear him down.’’ C
says “I’ve been torn down for-,” B says “Ok,” and C just stops when B starts
talking. Another one goes, A, talking to B says “And seven you’ve been
arrested for rape and other things. Ha.” C starts right after that with “Eight
you-,” B says “No, not that.” C stops, and then starts over, after “No, not
that” with “Eight you- Eight you give lip back to everybody.” 2 And if that
‘interrupter’ situation operates, it’s one way of seeing that you have one of
these first speaker rules that provides for a second speaker to talk, and the sort
of things he ought to do.

For the second speaker rules, one way to see if you’ve got something is to
see whether you’ve got an ordinary grammatical rule operating within a
sentence. One of the most prominent of the second speaker rules that you’ll
find in our data involves uses of ‘that,’ where ‘that’ seems to refer back to
whole clauses. “If you go hotrodding you’re bound to get caught” followed
by “That’s the problem with society.” On this class of phenomena there’s
very little literature. About the only thing I can offer deals with a very special
part of them, and those are things that could be called ‘sentence connectors’
or ‘clause connectors,’ which would go immediately in between two clauses,
typically where the clauses could be whole sentences. Things like ‘and,’ ‘that,’
etc. There’s a kind of a review article of sorts called ‘English sentence
connectors,’ by Seymour Chapman in A. Marquart (ed . ) , Studies in Language
and Literature (1954), page. 315.

1 See Spring 1966, lecture 7, pp. 323-4.

2 It seems that Sacks was working from his original 1965 transcript and not the
one attached to the Spring 1966 lectures. Here are those two fragments from the
original Sacks transcript:

Mel : Now let’s see what else can we decide about you?

Henry : Hey don’t tear him down

Joe : I've been tom down for-

Mel : ok

Henry : We got company

Henry : And seven you’ve been arrested for rape and other things. Ha

Mel : Eight you-

Bob : No. not that.

Mel : Eight you Eight you give lip back to everybody.

Lecture 15

379

Now, as I say, you can build hundreds of them if you like. The question
is, what do they do - besides providing for, in part, the non-disorderability of
the sequence of utterances? I take it that their fundamental use is that they
constitute a relatively formal apparatus for a ‘local check’ - by which I mean
at any given point - on the fact that what it is that’s just been said, is
understood. Now that’s tremendously important. What we have is a
machinery that’s built in, that holds for any conversation and for anything
that might be being talked about, that provides a demonstration and provides
for work that will demonstrate, that a speaker understands what a last speaker
says. Where such a demonstration does not involve time out from the
conversation, i.e., building a meta-conversation like: “We were in an
automobile discussion,” then, “I take it by an automobile discussion you
mean. . which is a theoretically possible way that one might have tests of
understanding built.

With these things, let me work out some of the ways that the rules get used
from the participants’ point of view, so that we can see what’s involved in
their use. Taking a typical situation: In some utterance of A, some noun or
some verb is used. In the next utterance, of B, one gets, not the noun or the
verb, but, say, a pronoun or pro-verb. Now that involves, for understanding
the second utterance, that one look, not, now, to what the term in it refers to
in the world, directly, but that one goes back into the conversation to see what
object in the conversation it refers to, and then see what that refers to. Thus,
any second speaker, using one of these rules, first builds his utterance by
performing some operation on some term of the preceding speaker, thereby
providing that he has some idea about what the preceding speaker is doing;
and now for the preceding speaker to understand the second utterance, the
preceding speaker has to do the same job that the second speaker did. That
is to say, he, himself has to see what that fellow means by ‘it,’ and find that,
‘By “it” he means the thing that I referred to by “automobile discussion” or
“hotrodding” or whatever else.’ And then he can proceed. So we have a rather
elaborate way that the two of them are checking each other out, and doing,
in their conversation, ways of talking that require that those operations be
done on prior talk.

So each part, then, of a conversation, becomes a piece of collaboration. One
isn’t getting a situation where, simply, A follows some rule and B follows
some rule, but whatever rule A follows, B has now to inspect to see what rule
he should follow, and that he followed it is something that A has now to
check out to see what he should do, etc., etc.

Certain relatively trivial but kind of nice consequences of the use of this
apparatus are that it makes a special task if you want to play dumb about
what another person is saying. It can be quite hard to do it in a decent way.
Quite hard because, for one, I’d like to see somebody try to systematically not
use these rules in building their own utterance - let’s say just the second
speaker rules. Let’s say you conform to all the first speaker rules as a second
speaker. Just try, as a second speaker, to not use second speaker rules like the
pronouns, the pro-verbs, the various things like ‘that,’ etc., etc. I think it’s

380

Part 111

extremely difficult to do, and once you use the pro-term, you give away your
understanding of what they’re talking about, even if you want to play dumb.

For example, we have an extended attempt to ‘play dumb’ in a
conversation that goes something like this. The situation is approximately the
following. A guy and a girl are driving around. They go over to somebody’s
house, a friend of the guy’s. He’s not home, but they meet the mother. He’s
away, having some troubles taken care of. They come out of the house. The
girl says to the guy, “Tell me about those people.” He doesn’t want to accept
that, so he says, “What do you mean, ‘tell you about them?” But the game
is all over by then, since, that he knows what she’s talking about is clear
enough given that he’s performed the operation on ‘those people’ to ‘them,’
where the sneaky part of the request was that there wasn’t any ‘those people’
except by virtue of the fact that the mother is seen as a representative of the
family, the family being the cause of the trouble, and known as such. What
he really would have to say, to play completely dumb successfully, would be,
“Who? Tell you about who? What do you mean ‘those people?” But she
knew, and he knew, that the trouble was properly explained by the way the
family behaves; that they got him into this. And in that case, you can see a
single person, ‘the mother,’ as one of ‘them.’

These tying terms provide, then, this constant local control on the fact that
somebody understands what somebody else is saying. Each person being able
to check it out, each person being forced to do the work constantly of
performing operations on the statement of another, and then performing the
very operations on his own previous statements, etc., etc.

Now there are lots of other sorts of tying objects. There are just ‘tying
terms’ themselves. There are a list of ‘tying verbs’ which, when they occur, are
then recognizable as second speaker uses. In this conversation, for example,
you’ll get somebody beginning their statement with “I agree.” Then there are
‘listers;’ for example, in the series of questions, ‘first,’ ‘second,’ etc. So there’s
a great raft of those sorts of things that lock conversation together. Our
problem is, at this point I could go through 35 of this rule, but I find it a
tremendous drag to do. Let me offer one that is interesting, anyway. It’s at
least not so completely obvious. I won’t give it abstractly, I’ll give a perfectly
concrete example.

A has a gun and is considering killing herself. B wants to know how come
she has the gun, where did she get it. So he says, “Is it yours?” She says “No,
it’s my husband’s.” Now that sort of thing - and it’s very generally used - I
call a ‘ correction— invitation device.’ And the way it works is, we might imagine
that a perfectly okay answer to “Is it yours?” is “No.” Routinely for a certain
class of questions, what one gets is not just ‘no’ or ‘yes,’ but ‘no’ plus - not
simply the right answer, but an answer that stands in a strong relationship to
the item used in the question. Where one sets up a special sort of problem by
the item one uses in the question. It can be, for example, a request for a good
reason why something is held.

For example, later on in that conversation B wants to know why A has the
gun. He says “Is he a police officer?,” where ‘police officer’ is now a good

Lecture 15

381

account for somebody having a gun, and is to be replaced by, not simply “No
he isn’t,” or “No he isn’t” plus “He’s a butcher,” but “No he isn’t” plus
‘why we have the gun.’ Where you wouldn’t get, then, his occupation, if he
isn’t somebody other than a police officer who has a right to have a gun, but
some other account of why the gun is had.

The item, then, in the ‘correction-invitation’ thing, is used as a ‘possibly
good account’ for which, if it’s not the answer, give me a correct good answer
if you have one. Where you suggest what the sort of thing is that would be
‘a good answer.’ And in those cases, then, the item offered with ‘no,’ as the
answer to the question, is tied, not simply to the item in the question, but it’s
tied via some much stronger kinds of constraints, such as ‘co-class member-
ship’ of a sort, with the item in the question.

One other thing about the first speaker rules. Another way of seeing that
you have first speaker rules concerns an issue which I mentioned earlier, and
that is the special possibility of conversations of a form, not A-B-A-B, but
A-A-B, for example, and that you could find those, and the way they tend to
be found in the easiest fashion is where some first speaker rule is used, and no
use of it by a second speaker is made - he doesn’t say anything, for example
or in any event he doesn’t say what would be a return to that first speaker rule.
Then what one gets is a repeat, where the repeat is not simply an extension
of A, but A was complete, and now what we’ve got is A-A. And not simply
a ‘repeat,’ but an indication that no B has occurred and that B ought to have
occurred, i.e., that a first speaker rule was used. Trivially: “Hi.” No answer.
“Hi." We would write that, not as one long A, but given that the first speaker
rule items tend to be adequate complete utterances, on their completion we
have the B. If we don’t get it, and we get A repeated, we have A-A. Those
are possible formats for conversation at certain points. And where they are
operative, we can say that we have ‘a first speaker rule item again,’ where we
don’t have a way of otherwise saying that.

Lecture 16

Possessive pronouns ; Possessables and
possessistives

I’m going to begin talking about this phrase, “In that Bonneville of mine”,
first partially focussing on the term ‘mine’ in it. ‘Mine’ is of course one term
of the so-called class, ‘possessive pronouns.’ I say so-called because the literal
use of the notion that they are possessive pronouns made considerable trouble
for more than a few analyses of the problem of possessions. Rather regularly,
at least in the past, attemps have been made to formulate the notions of
ownership in some tribe by considering the types, and range of use, of
possessive pronouns.

A rather classic example, referring to earlier ones as well, can be found in
Volume 5 of the Southwestern Journal of Anthropology, pages 169-189, an
article by Capell entitled ‘The concept of ownership in the languages of
Australia and the Pacific’. What we have there is an exploration of, and an
attempt to explain, the use of possessive pronouns with respect to ownable
objects, and to derive what he calls ‘psychological notions’ of the differences
in conception of what can be owned, how it can be owned, etc., as a way of
accounting for the differences in the use of the possessive terms - possessive
suffixes, possessive prefixes - for various objects.

That conception has been criticized somewhat justly by Goodenough (I say
‘somewhat’ not that the criticism is only in part just, but because the concept
can be subject to much more serious criticism) in a very famous book entitled
Property, Kin, and Community on Truk, Yale University Publications in
Anthropology, no. 46.

Now Goodenough proposes that you have to determine conceptions of
property independently of the possessive pronoun uses. And he offers some
criteria for something being a possession which are not directly usable for our
purposes, but I’ll touch on them in a bit. What we want is to find out what
it is that provides for the use of ‘mine’ - both in this phrase, and for such uses
for other objects by other persons. I said “so-called possessive pronouns”, and
I said that the attention was over-literal to the name, because it’s perfectly
plain for English, and it’s probably perfectly plain for most of the other
languages about which such analyses have been done, that possessive
pronouns are used routinely in ways which only absurdly can be considered to
be ‘possessive,’ i.e., in the sense that one could say “I own that car.” Before,
however, we would say that we had a set of homonyms, only incidently

Includes pp. 1-7 of the original Spring 1966, lecture 17

382

Lecture 16

383

having the same sound between, say, ‘my’ in “my car’’ and ‘my’ in “my
country’’ or in “my brother’’, we have to see whether there isn’t some
important relationship between some of those terms and some of the different
uses of them.

We could start by separating out some of the uses. So, for example, there
is - in English anyway - a fairly clear ‘affiliative’ use of the possessive
pronouns: When one has a membership categorization device category, to
which one can attach a possessive pronoun, then what one is doing is claiming
membership, asserting membership, in that category, or in a case of that
category. “My family,’’ “my country”, and the like.

A second, sometimes related use, seems heavily to involve the set of
relational pairs of persons, where if there is some relational pair between A
and B, then either A or B can refer to the other by reference to one of the
possessive pronouns, particularly ‘my.’ “My brother”, “my doctor”, “my
patient”, “my boss”, “my child”, etc. And if the warrant of the affiliatives is
that one is a member, then the warrant of one of these is that there is such a
pair, with incumbents, and that one is in such a pair.

Then of course there is the ‘possessional’ use of the possessive pronouns,
with respect to, obviously, objects of possession. “My car”, “my hat”. And
I’m sticking to ‘my’ for now, because the use of ‘my’ and the use of ‘your’
seem perhaps to be regulated in quite different ways.

What we then arrive at is: It is not the use of a possessive pronoun that tells
you that the object it is used with respect to is in some way owned, but it’s
by independent determination of the character of that object referred to, that
a possessional use is heard. That is, it is by virtue of the fact that something
is, first, roughly a possession - and not, for example, a membership
categorization device category or a relational pair category - that tells you,
when a possessive pronoun is being used, that it’s being used to note, claim,
force, whatever, a possessional relation.

That then poses the task of determining what objects are possessions,
recognizably so. And that being the case, we can consider Goodenough’s
criteria. One tendency of his in constructing criteria for deciding that
something is a possession is to consider such things as: What use does the
object have? Objects used as food are to be seen as possessions. How is it that
the thing is produced? Objects produced by man’s labor are possessions. One
can have property rights, etc. Criteria like that.

There’s no doubt that there’s some sense in which those criteria have some
real relavance, but there are large problems with those criteria. Our concern
is regularly with recognition problems. And what we want are criteria that
deal with recognition problems. I want to make a couple of distinctions,
because of the character of the term ‘possession’ which seems necessarily to
suggest that the thing is known to be owned by somebody or some group or
whatever, and where we take it that it’s individual things that are possessions.

I want to distinguish, now, two things. One, I’m going to call ‘ possessables.'
And that has its usual lay sense, i.e., these are things which you could, if you
want, own. The other is a rather central class, which I’ll just give a slight

384

Part III

twist-name to, ‘ possessitives - it doesn’t have any intuitive meaning. My big
concern is with ‘possessitives’ because they’re the things that are crucial to the
recognition problem. Consider the following event: Somebody’s in a room,
they pick up a hat, and they say “Who’s hat is this?” How are we going to
deal with the fact that they see the hat and, never having seen the hat before
perhaps, and not knowing whose it is, and therefore in a sense now knowing
that it’s anybody’s, they know that it’s somebody’s.

‘Possessitives’ are a class of classes of objects which, when cases of the class
are encountered or talked of, they’re recognized to be somebody’s possession.

Criteria for membership in the class ‘possessidve’ can be worked out in
fairly nice ways. And the criteria are probably not local to this culture. In a
fairly offhand way, Goodenough, in his treatment of possible ways one can
come to have possessions, suggests something that the analysis I’ll develop
could account for. Let me quote his remarks and then we’ll take off from
there. This is from page 35 of Property, Kin, and Community on Truk :

If a man discovers goods which another has lost, he acquires full title to
them, provided the loser is not a person who is known to him (i.e., a
member of the same community or kinsman from another island). The
discoverer is owner because he brings the goods into existence as
property within the community of persons with whom he interacts. If
the former owner is known to the finder, however, their relationship to
each other is presumed to have been established already with respect to
the goods lost. And the finder must return them or be branded a thief.

On the other hand, if the loser is a stranger from another island, whose
lost canoe, let us say, has drifted over to the finder’s island, then the loser
must buy it back from the finder just as though he were making a new
purchase.

Now there’s a kind of nice thing involved there. And it’s nice in that it will
give us some rather important socialization problems, which we well know to
exist. And that is, that we can distinguish between things that can be said to
be ‘found’ - and such things are ‘possessables’ - and things which, having
found them, you find them as things that somebody has lost, i.e., where what
you see when you find it that somebody lost it. Again, now, it’s an item
you've never seen before, necessarily. That is, you will know the class, but it’s
not by virtue of the fact that you see ‘Johnny’s hat’ that you know that it’s
the hat that Johnny lost. (Indeed, even under that circumstance it’s an
interesting issue that, knowing it’s Johnny’s hat, you know that where you
found it, you found it on the occasion of his having lost it, rather than, for
example, that he threw it away. That is, you don’t have to know that Johnny
has put up a notice, ‘I lost my hat,’ to see that the hat you found is the hat
he lost.)

So we’re saying right off that ‘possessitives’ are things which, for example,
when found are seen as having been lost, i.e., are seen as having been owned
by somebody. That being the case, we can get an initial list of some of them

Lecture 16

385

anyway, by going to Lost and Founds and seeing what gets turned up. Not
everything that’s found is seen as something that’s lost. If you go down to the
beach, pick up a stone, you can make it your own, it can become a possession.
But I take it that if you turned the stone in at the Lost and Found, they might
take it that something strange is involved.

And there are a variety of other properties that ‘possessitives’ have. Some
of them can be used to acquire other property, i.e., they’re generative. You
claim something by putting a possessitive on it. You put a book on a table in
a cafeteria, people take it that somebody’s using that table. You can play
games by seeing what sorts of objects you can put down there that do not
interfere in other people’s sitting at that table, and come to a differentiation
at boundaries of the two classes.

That then gives us further kinds of considerations; for example, if you have
a possessable that’s not a ‘possessitive,’ then you have to be very careful about
treating it so as to retain it, whereas if you have a possessitive you can be much
more casual about it - drop it in the street, leave it in front of your house -
and people will not thereby take it that if they want it they can just pick it up.
Or, if they do, then you can well claim that they’re thieves. You can leave a
car in the street, it may be robbed, but you needn’t fear that someone will
figure by virtue of the fact that it’s sitting on the street that it’s available for
whomsoever wants it. And with cars, of course, having their locus in the
street, there is then the special problem of detecting that they’re abandoned.
Cars can sit for enormous long times on the street, intending to have been
abandoned and not recognized as such. And there are classical stories of
people attempting to get rid of an object, finding that it invariably pursued
them unless they found an extremely good place to get rid of it, as it was
regularly returned. Where, then, there are special places to leave things which
permit others to see that they have been abandoned.

So you have a very strong sense of a difference. And that differences makes
a nice discrimination in certain occupations. There are two related occupations
- begging and scavenging - which involve quite different obligations and
quite different activities. Beggars ask for things. For the things they acquire,
they have to go through some operation of getting by asking. Scavengers are
a much more independent sort. They go to those places where people
abandon things, and pick them up. And in doing so they take on no
obligation of thanking, or whatever else, the people from whom they acquire
the objects.

Again, then, we do not get a list of particular objects making up the
‘possessitives.’ And there’s no need for Members to have encountered that
particular object before, nor for them to know whose it is, to know that it’s
somebody’s possession. Though it’s conceivable - but I doubt it - that on a
very small island everything would be known in that fashion, i.e., one could
always say about something one encounters, “That’s so-and-so’s such-and-
such’’. And it will certainly be the case that a great many possessitives are
understood to be possessitives by virtue of how they’re produced, or by virtue
of their uses. But that’s an independent matter that controls, for one, how it

386

Part III

is that Members come to put a class of objects into the class ‘possessitives.’
Now, I mentioned that the differentiation between the two classes,
‘possessitives’ and ‘possessables,’ gives us some rather important socialization
tasks. That is, coming to see ‘possessitives’ and coming to know how it is that
a possessitive can be acquired, free, has got to be learned, that is perfectly
plain. And it’s quite extraordinary how early that is learned in this society. I
can show that by a little piece of data which I’ll quote. The data is from a
book entitled One Boy’s Day by Barker and Wright (1951), pages 379-380.
What they did was to have a bunch of people, working in half-hour shifts,
follow a kid around all day, writing down as much as they could of what he
did. They compiled, then, a record of his day. The boy is, I think, seven years
old. He’s been playing with a crate which is in a lot at which there was a house
that burned down, though the lot is still owned by the owner of the house.
The family is having dinner, and the father is talking about the crate:

F: What are you going to do with that big crate you have?

S: I’m going to bring it home.

F: You are? Don’t you think we have enough crates and that kind of

stuff around here?

The son doesn’t give an answer. The mother then talks:

M : What are you going to do? Use it for a tent?

V: Yes.

M : Maybe Mr Nelson wanted it. Did you ever think of that?

S : No. He was going to burn it. He doesn’t want it.

As you see, not “No I didn’t think of that.’’

F : How do you know? Did you ask him?

S: No.

F : Are you sure?

S : Yes, I’m sure.

M : Was it there where they usually burn things, on that rubbish heap?

■S': Yes.

M : How are you going to get it here?

S : I’m going to put it on my wagon and bring it over.

As you can see, the parents know how it is that the thing could be recognized
as having been abandoned. Now, there are a lot of things involved; for one,
certain intentions are taken to be ‘intentions to give it up’. That is, it’s not, if
Mr Nelson wants to burn the crate, and puts it on the rubbish heap, he can
then say “Well, I really wanted to burn it, I didn’t want anybody who wanted
it to take it.” He seemingly has given up some of his rights over it. But the
parents take it that the kid could have known how it is that this thing could
be intended to be disposed of, i.e., that it was placed on the lot, and that the

Lecture 16

387

kid can inspect the lot to see what things are available for anybody, and
acquire them in that manner. But that if he hasn’t acquired them in that
manner, then he has wrongly acquired them - or wrongly intended to acquire
them. And it’s not necessary, apparently, to ask what the intention about the
object is; if it’s properly placed, the intention is available through the
positioning.

There are, then, a large class of objects which when seen, or when talked
of, are recognized as somebody’s, or some unit’s, and whose they are need not
be known to know that. But we presumably don’t need the distinction
between ‘possessitive’ and ‘possessable’ as a notion, for the issue of how it is
that possessional reference, in general, is heard. That is, when either a
possessitive or a possessable is used in combination with a possessive pronoun,
we can take it that that sense of the object as ‘a possession’ is what’s being
referred to. So, for example, when we get a phrase like “that Bonneville of
mine”, we have no trouble in determining that it is possessional reference that
is being made. And of course we can’t do that by such a fact as, it’s a noun
being used with a possessive pronoun. If we use that criterion - noun/
possessive pronoun - we get equally ‘that brother of mine’, and the like. And
those are not the same sorts of things. No possessional reference is indicated.

Now I’ve gone through some of the ways that the possessive pronouns are
used: An ‘affiliative’ use, in respect to some category of a membership
categorization device, a ‘relational’ use, with respect to some relational pair, or
category of relational pairs, and the ‘possessional’ use. And I’ve argued that
one independently determines that some object is a possessitive or possessable
so as to see what use the possessive pronoun has, and if the object being talked
of is a possessitive or possessable, then one can hear the possessive pronoun
being used possessionally.

But that’s only to say that if we have that sort of object being referred to,
then some possessive pronoun can be used possessionally - not which one, or
how which ones are selected. There are a variety of ways and I want to
consider a couple, one of which will give us a relationshp between several of
the types of uses of possessive pronouns - most particularly between the
usability of them affiliatively and the usability of them possessionally.

And that matter is fairly simple, at least to some extent. If, for some
membership categorization device category, one can use a possessive pronoun
affiliatively - one can say, for example, “my country”, “my family”, and the
like - then, if there is some object which can be a possession for some class of
those membership categories, then one can use at least the first-person plural
possessive pronoun for that object, possessively. That is to say, if one can say
“my country”, one can say about something that’s a possession of that
country, at least ‘ours.’ “My country,” “our bases;” “my family,” “our
house.” (It isn’t the case that one can automatically say ‘my’ for such an
object.)

One can’t routinely do that for objects which are seen as possessions of a
correlative member of one of the relational pairs. So the fact that you can use
a possessive pronoun for some category of a relational pair does not permit

388

Part III

you to use a possessive pronoun possessively with respect to something that is
owned by the pair member, or is seen as a possession of the pair member. You
wouldn’t, for example, say, if it’s my doctor’s chair’ that it’s ‘mine’ or ‘ours.’

Now one of the things we can see when we have some unit like ‘country’
or ‘family’ for which the object is a possession and we get this rule that
permits one to use ‘our’ for it, is that it’s routinely the case that that use of
‘our’ is one that a lot of people other than oneself can do, i.e., there are a
whole range of people who can say about those bases that they are ‘our bases’.
When we get a use of ‘my’ for some object, what we have to decide is how
that use of ‘my’ is arrived at, and when we get how that use of ‘my’ is arrived
at, we can have some way of deciding whether there are others who can also
use that ‘my’ about the same object. So, for example, if you hear some guy
using ‘my car,’ and it turns out he is talking about the family car (where he
can use that by virtue of the fact that if it’s the family car he can at least say
‘our’ and perhaps ‘my‘), then it may also be the case that his brother can say,
about the same car, ‘my car.’ So there may be a set of people who can say ‘my’
just as well.

That has, however, to be worked out for the sorts of objects, and the ways
that those sorts of objects are seen as possessed. And the question of what
controls the use of ‘mine’ is, in part, not resolvable by reference to, for
example, a search of the legal title. It is, for a class of objects, to some
considerable extent a problem of address. That is to say, it’s a problem of
who’s being talked to. We get a tremendous instability in what possessive
pronoun is used for a given object over the various pairs of persons who may
be talking, so that ‘the very same car’ talked about by A, can to one person
be referred to as ‘our car,’ to another be referred to as ‘your car,’ and to a third
be referred to as 'my car.’ That’s all perfectly possible - and possible,
furthermore, under a situation where no one of those people have title to the
car.

It is often the case that the sense in which ‘mine’ is used, i.e., what other
terms would be used in other environments, can be gotten by a hearer from
the sort of object referred to by it. That is to say, if I say “my toothbrush,”
you who hear it can be fairly assured that there isn’t any other environment
under which somebody else would refer to that toothbrush as ‘mine,’ or I
would refer to it as ‘ours’ or ‘yours,’ whereas for other objects you can feel
assured sometimes that depending upon who’s speaking, there will be
environments when they can’t say ‘mine.’ And that fact can be quite crucial.
That is to say, if a kid says to another kid “my Bonneville,” the other kid may
well be able to know that the Bonneville he is talking about as ‘mine’ he can
talk about as ‘mine’ talking to me, but talking to his parents, for example, he
can’t. And the relevance of that, roughly at this point is: Insofar as a car can
be used to see the independent status of somebody, then if you know that they
can’t use ‘my’ for all comers about it, then you can feel that they’re not
independent for that item. That is, if someone says “that Bonneville of mine,”
somebody who doesn’t know them can see, and say, “It’s the car your daddy
bought you.”

Lecture 17

Pervasive, inexhaustable topics;
Emblems

I’m going to continue talking about the phrase, “In that Bonneville of mine.”
There are a large series of issues, really quite central to this piece of data. I’ll
start off with a quote from chapter 1 of The Nuer by E. E. Evans-Pritchard.
In that chapter he makes a whole set of observations parallel to the ones I’ve
been making, and much related to the stuff we’ve been doing on the
phenomenon of an ‘automobile discussion’ among teenagers. The quote is
from pages 18-19.

We have seen in a brief survey of some Nuer institutions and customs
that most of their social behavior directly concerns their cattle. A fuller
study of their culture would show everywhere the same dominant
interest in cattle; for example, in their folklore. They are always talking
about their beasts. I used to sometimes despair that I never discussed
anything with the young men but livestock and girls. And even the
subject of girls led inevitably to that of cattle. Start on whatever subject
I would, and approach it from whatever angle, we would soon be
speaking of cows and oxen, heifers and steers, rams and sheep, he-goats
and she-goats, calves, and lambs, and kids. I’ve already indicated that
this obsession - for such it seems to the outsider - is due not only to the
great economic value of cattle, but also to the fact that they are linked
in numerous social relationships. Nuer tend to define all social processes
and relationships in terms of cattle. Their social idiom is a bovine idiom.

Consequently, he who lives among Nuer and wishes to understand
their social life, must first master a vocabulary referring to cattle, and to
the life of the herds. Such complicated discussions as those which take
place in negotiations of marriage, in ritual situations, and in legal
disputes, can only be followed when one understands the difficult cattle
terminology of colours, ages, sexes, and so forth.

Now there’s a big problem with these observations. It’s one thing to say
that every conversation ends up about cattle, and another to show how that’s
so. And the point about a central conversational phenomenon and its analysis

This version of lecture 1 7 begins with a consideration first given at the end of the
original Spring 1966, lecture 16 (pp. 16-17) and more fully developed in Fall 19 66,
lecture 5, from which the bulk of the discussion shown here is taken.

389

390

Part III

- for example, that it’s due to everything being tied relationally to cattle - is
obscure in that it supposes that in the end it would be some formulation of
social relationships - economics, marriage, etc. - that is fundamental to the
culture, so that they provide the sort of inevitable explanation one would
come to. That is a very important theoretical issue. And that it is solved by
him in this fashion - and by most of sociology and anthropology in a like
fashion — is the sort of thing that permits him to say ‘it would turn out that
. . .’ when it hasn’t yet been investigated.

That is, one sees that something - like cattle conversation for Nuer - is
important, it seems extremely pervasive. And now, before trying to figure out
how and why it works as it does, one uses an acceptable explanation as that
which the analysis would eventually and inevitably reveal. By ‘acceptable
explanation’ I mean one need only invoke what’s known to be such, e.g.,
‘social relationships,’ as the stable machinery generating conversation, etc.
And insofar as investigations end at the point when the explanation is in sight
but hasn’t been developed, then the notion that those things are the ultimate
explainers can remain in use, and seem even to be confirmed. Whereas it’s at
least theoretically conceivable that, having examined the structure and
working of these cattle conversations, one consequence might have been a
discovery that the supposed explanation didn’t do the job. We can’t suppose
that some end, being in sight that way, would inevitably be arrived at.

Evans-Pritchard does, however, point up that a topic can have a really
pervasive character for a culture. Now the issue is not, then, to say that
perhaps for kids, automobile discussions are like cattle conversations are for
the Nuer. But perhaps there’s a phenomenon - pervasively important
conversational topics - which is present in both. Where what would perhaps
be of interest would be that there wasn’t such a topic.

One thing, which is not quite the same as the kind of point that
Evans-Pritchard makes, is that these are topics which can be called ‘ultra-
rich.’ There are different ways in which something can be ultra-rich as a topic.
In the case of an automobile discussion, you can come on one of the ways by
the following line of reasoning. You could say that a topic is ‘limited’ if there
were, say, a finite set of things that could be said about it, in such a way that,
as you were talking about it you were cutting into the amount that could be
said, and thereby reducing what could be said some other time, so that at
some point the topic is exhausted. By the term ‘ultra-rich,’ one thing we mean
is that the topic is inexhaustable.

Now, saying that it’s inexhaustable and finding ways of showing that it’s
inexhaustable are altogether different things. And in making the remarks I’m
going to make now about its inexhaustability, I’m using other data than our
1 1-page segment, but in fact I’m using limited data, and limited data that I
think will serve. We might go about the task of finding ‘inexhaustability’
along the following lines. Suppose we could say there were a number of finite
parts to a car. And there are rather obvious ways in which you could say that,
turning on the lists of parts put into a car. Now, if we imagine that, in part,
an automobile discussion consisted of talk about the parts, and that, say, for

Lecture 1 7

391

any given part there was only so much that could be said, then it would be
at least arguable that, the parts being finite, the discussion of an automobile
is, in that sense, finite.

There are two interesting properties of automobiles for hotrodders, which
undercut such a line of reasoning, and provide a way of seeing the
inexhaustability, i.e., the infiniteness of an automobile discussion for them -
using now, a very conventional notion of ‘infinite.’ First of all, any part,
perhaps, but certainly many parts, of any car whatever shape it’s in, say,
even right off the assembly line, is replaceable. And the second fact is that
any replaced part is replaceable. That means that any car now - looking to
the future - being considered, consists of an infinite set of parts, i.e., any
part it has, plus its possible replacements. And all that we need, to talk
about something being ‘infinite’, is that for any finite set, another can be
added.

And that is quite different than the way adults talk about cars, and the way
adults conceive cars. So if I, talking as an adult, were to say that I just replaced
the engine or the tires or something else on my car, you might figure that there
was something wrong with it. And that having replaced it, it was over. For
kids, that’s not so. Any part is replaceable without regard to the fact that it’s
in perfect shape, by virtue of the consideration, e.g., that a better part has
become available, or by virtue of the current combination of parts that I have,
some part is better than another part that I have, where all that is generative
in the sense that the aim is to have ‘the best car.’ And the notion of that as
a ‘central motive’ is something that we could work out eventually.

With 'having the best car’ as a ‘central motive’, then if any set of cars that
you’re considering as equivalent, or possibly faster, get modified by having a
new part put in, then that part is eligible for consideration in your car. You’ll
find in kids’ conversations about cars, that the issue of indefinite replaceability
of any part is routinely present, and it’s present without the characteristics that
replaceability has for adults. If a kid proposes to put a new engine in a car,
the issue is not whether the old one was worn out. It’s merely that a ‘better
one’ has become available, or the kid can afford a better one, or someday will
be able to afford a better one. In that sense, then, the purchase of any car is
only the purchase of an object which is then to be considered in each of its
parts, for indefinite replacement. And that is altogether different than the way
adults properly conceive the cars that they buy. Where, furthermore, for the
kid, each actual replacement is a matter that is discussable, and any definite
positive decision is relevant for all participants. So that if A comes up with a
good reason for putting 20-inch slicks on his car, I have to consider them for
mine.

Furthermore, it’s not any population of kids who know each other that
constitutes the population for these replacements, considerations, and assess-
ments, but it’s perfectly legitimate to propose as a reason for replacement that
you heard that somebody did it, or that you read that it’s a good thing to do.
And in that sense, then, the whole population of hotrodders is a group whose
actions are relevant for any member of that population. And under that

392

Part III

formulation of the discussability of parts, with the parts having this infinite
character, the topic can be seen to be inexhaustable.

It further appears that Evans-Pritchard’s observation that almost any topic
among the Nuer is one that eventually can come around to being discussed
by reference to cattle is relevant here. Recall my earlier remark that there is a
very large range of topics conventionally separated from automobiles that are
discussable via automobiles - independence, sex, relations with the cops, how
to prove yourself, comparability of persons, the state of the modern world,
etc., which made it as ideal a topic as could be had, to serve for the members
of this group as a ‘cover topic. ’ 1

Let me turn now to some issues by reference to, more specifically, the
phrase “In that Bonneville of mine.” For one, there are questions of the
relevancies of alternative names that can be given to the same car. The
‘Bonneville’ is subsequently talked about as a ‘Pontiac station wagon.’ Earlier,
that sort of car is referred to as a ‘Daddy’s car:’ “Now the Soshes keep
driving. 2 Daddy’s car and everything” (data handout, first page). And the
user of ‘Bonneville’ recognizes it, furthermore, as looking like “a damn
Mommy’s car.” And now there could be a series of other types of references,
each of which may have different sorts of relevancy for a variety of things.
One, for example, is whether you can go drag racing in it. The speaker here
takes it that you can. His colleagues, as I’ll try to dig out later on, are rather
insistent that you can't.

There’s another set of alternatives which concern, whatever names you
might give to this car, its position as a type of car among alternative types of
cars: Bonne villes as contrasted with Chevvies. That kind of car contrasted
with hotrods. And that set of alternatives has another class of relevancies. I’ll
just briefly talk about those, and I’ll begin by talking about a class of objects
that quite obviously have the properties I want to consider.

There are a bunch of things we can talk about as 'emblems’ . These are
things like medals, titles, and the like. They have several important
properties. First, if you see one, see somebody wearing one, hear somebody
addressed by one, and the like, then you can know the sort of procedure
whereby they came to have it. And that you know the sort of procedure
whereby they came to have it is crucial to the assessment you make of it. For
example, that it’s something they should be proud of or embarrassed about
- proud of, in the case, say, of a medal; embarrassed about, classically, say,
the mark of Cain, the clipping of ex-prisoners’ ears, and things like that.

The availability of those procedures whereby they came to get them is, of
course, independent of particular information as to how this fellow got that
thing. One can know how people get such things, and be assured that not
only do people get such things in such a way, but this fellow got that thing
in such a way. That knowledge is, to use a phrase I introduced earlier,
‘protected against induction.’ And that it is, can be seen from the following

1 See Spring 1966, lecture 7, p. 320.

2 See fn 1, p. 396.

Lecture 17

393

sort of colloquy: A says, “What did you do to get that medal?” B says,
“Nothing.” A says, “Oh you’re being modest.” In order to be seen as ‘being
modest,’ one has to take it that what B replies with is not treatable as possibly
undercutting one’s knowledge of how it is that those things are acquired.

One wants to note, not simply in passing, that the fact that when one sees
emblems, one sees them as having been properly acquired, can make for a
little trouble for arguments about things like possessions having their
character for Members as ‘possessions’ by virtue of Members knowing the
way that they’re acquired. That is to say, one doesn’t know that that car is a
possession because one knows that it was purchased; one sees it and sees that
it is a possession. One sees, as well, that it was properly acquired, and the way
that it is properly acquired is by being bought, and one doesn’t have to check
that out. Indeed, if one does have to check it out, there’s trouble.

The fact that emblems are seen as properly acquired poses a series of
possibilities which are, in fact, actualized regularly. Two correlative ones are:
First, it’s an opportunity for somebody to acquire the thing how they may,
knowing that it will be seen as having been properly acquired. Having it is
proof that you acquired it properly, and gets whatever might be gotten if you
did properly acquire it. That clearly enough is a case of subversively using its
properties. One can buy, and people do buy, medals in pawn shops, titles
from mail-order universities, etc.

Now there’s a negative aspect to the very same thing, and that is, for those
things which, having acquired them, Members know how you acquired them,
there have regularly been things that people can’t, therefore, acquire. So, for
example, it’s important to, say, the development of modern capitalism that it
became the case that one couldn’t tell about somebody who has money, how
they acquired it. Until that was so, there were large bodies of people - the
aristocracy - who couldn’t go into business and get money because it would
be available that they were ‘in trade’ - something bad - if they had money.

The second correlate of this availability of how it is that something is
acquired, is that if you want to go into the business of debunking, then, for
any object which people know, and sanctionably know, how it is that such a
thing is acquired, one need only seek out cases of its subversive acquisition and
write them up. That’s always of interest, and there is no shortage of books and
movies on how it is that people really got Nobel prizes and the like. You can
arrive at those sorts of things quite formally - and that provides an absolute
set up for, e.g., having computers write scandalous literature.

Now there are lots of possessions that are very similar to emblems in this
sense that we’ve talked about emblems. How is it that they’re acquired is
regularly available to Members, and they can use their knowledge of how it
is acquired to see one as something you should be embarrassed about or proud
of. We want to consider the case of cars, but for many other possessions it
works in a similar fashion.

A given car, the car you have, is treated as a case of a set of possible cars,
and the set of them provides those things from among which the given one
was selected. Then, Members apparently know what different selection

394

Part III

procedures will come up with. They can then take the given car, construct or
invoke the selection procedure that would come up with it and not some
other, and see in a glance that that selection procedure provided for the
purchase of this particular car, where the selection procedure may have a
variety of further relevancies. We then need to consider, for one, what scope
does the set of alternatives have for any given object? That is, what things are
seen as possibly alternative to it so that they can be seen, perhaps, as having
been rejected? What those alternatives look like, I can’t really say. But I would
think, for example, if you’re wearing a beautiful ring, then somebody may
perfectly well see, not that you chose among the set of ‘rings,’ but you might
have chosen among the set of ‘jewelry,’ for example, and ‘necklaces’ might
well be considered as an alternative that you could have picked.

Now suppose that one criterion for the class - though it’s a kind of
equivocal one - would be that if you start with some other item than the one
in fact selected, you have available as an alternative to that other member, the
very one you selected, i.e., you can get back to the one you selected. So that
one can see ‘cars’ and ‘jewelry’ as not alternatives if, when you’re wearing a
ring, somebody doesn’t feel that they can see that you turned down a car to
get the ring. Again, this is without regard to how you in fact did it. You may
well have decided to go on a vacation instead of buying a car. That’s not the
issue. If you go on a vacation, people may well feel that what’s relevant is how
come you chose to go here and not elsewhere. And of course that’s very
important for trying to make it in certain groups. If you saved up all your
money to go on a vacation somewhere, then you may be able to pass as
somebody who chose to go here and not there, rather than somebody who
chose to go on a vacation rather than eat a month from now.

Let’s get to these known decision processes by reference to kids and cars. In
the first instance, whatever problems we might have constructing the
alternatives for, say, an adult woman, we don’t have for kids, since kids want
cars. If a kid has a car we can take it that he wanted a car, where the issue is
whether he wanted this one. And again, whether he wanted this is not a
problem of psychology; that is to say, it’s not a question of the sort that we
have to find out if he wanted this car by asking him. He can say that he
wanted the car, and you still know that he didn’t want it. You can decide that
he didn’t want it without regard to asking him.

So there are, then, the set of cars that are going to be used to decide how
this car was selected. There are a variety of formulations of car types, and
those car types are formulated by reference to different membership catego-
ries. That is, there are, for example, ‘Mommy’s cars’ and ‘Kids’ cars.’ And
again, the phenomenon of ‘Mommy’s cars’ is protected against induction. It
isn’t the case that if you think that a Pontiac station wagon is a Mommy’s car,
then if you see a kid driving one you say, “I guess I’m wrong.” You see ‘a kid
driving a Mommy’s car.’

(In that regard, one might consider the set of scientists’ sins, i.e., things
that, as a scientist, it would be a sin to do. In this case, that you retain a theory
over cases that are inconsistent with it, which a scientist doesn’t properly do.

Lecture 17

395

One might see whether, and to what extent, scientists’ sins are not inventions
of scientists, but involve the use of perfectly correct ways of reasoning, and
sanctioned ways of reasoning, when one is doing other tasks.)

If a kid’s going to get a car, then kids seem to know what sorts of cars a
kid ‘wants,’ i.e., a kid ought to want, it will be insisted he should want, and
he’d better want, otherwise he’s subject to being seen as not having gotten
what he wants, anyway. The others take it that they can see that that car was
purchased by Daddy, and that the kid who’s driving it only had partial say
over what car he got. He would find himself walking around with a constant
apology on his lips, i.e. ‘I really wanted this.’ It’s a sort of apology that can
be made for whole classes of things, in which your insistence is seen as
over-insistence - or wrong, in any event.

Lecture 18

‘Hotrodders’ as a revolutionary

category

Towards the end of the last consideration I was focussing on what kids take
it they can see, when they see some combination of a car and a driver; where,
as a rather minor point of what I wanted to get to, was that for certain cars,
driven by kids, kids take it that they can see that that car was purchased by
Daddy, and that the kid who’s driving it only had partial say over what car
he’s going to get. And it’s not only that a kid doesn’t want a car that’s going
to be seen as ‘a Daddy’s car’ or ‘a Mommy’s car,’ but that even if a kid wants
a Bonneville, then one of the things he has to consider is that it’s going to be
seen as ‘a Daddy’s car.’

Further, kids give some formulations of kids who drive Daddy’s cars; for
example, ‘Soshe,’ which is a term that kids use, by and large in a derogatory
manner. 1

But the problem is that these considerations don’t at all get at what’s most
important about these phenomena. And that’s what I want to try to do,
though the analysis I’ll offer has a much stronger sound than I could defend,
at this point anyway. But I’ll do it because something like what I’m saying is,
I think, involved, and it’s really fundamental to see.

We could work at it, in part, by asking a bunch of questions: Why do kids
go about making up all these typologies of cars - and the typologies of cars that
they have are really enormously elaborate: Daddy’s cars and Mommy’s cars,
gassers, turbines, dragsters, plus all the names of cars - Fords, Chevvies, etc.,
plus various modifications of those. And they use those typologies to make
assessments of other drivers; assessments which are not always nice. Why do
they do it? Aren’t the terms that are used before they go to work good enough?
And what’s the matter with them if they aren’t? So that’s the problem.

One way we can try to get at it is, in part, to try to see what kind of
difference there is between the category ‘teenager’ and the category ‘hotrod-
der.’ I want to argue that they’re fundamentally different types of categories;
that ‘hotrodder’ is, in a very non-trivial sense, quite a revolutionary kind of
category. And I’ll just be trying to sketch some of the ways that that’s so, and
offer others like it.

Because I take it that the issue is really a very deep one, I’ll give some

1 ‘Soshe (pronounced ‘sosh’) seems to have been a geographically and temporally
local derogative used by at least these teenagers for their more 'socially successful’
peers.

396

Lecture 18

397

suggestions about a type of category by taking a very ancient, and in its way
famous, example - though most people don’t know the example, so it’s of
some educational value, anyway. This is a quotation from Genesis 14. It goes
like this:

A fugitive brought the news to Abraham the Hebrew, who was
camping at the terebinths of Mamre the Amorite, kinsman of Eshkol
and Aner, these being confederates of Abraham.

In the history of biblical criticism, that’s a most important little section, and
its fundamental importance is at least sketched by the following kinds of
considerations.

The phrase ‘Abraham the Hebrew’ is apparently unique in the bible, in
that it’s pretty much the only place that an Israelite is referred to by the term
‘the Hebrew.’ That term, ‘the Hebrew,’ is only used by an Israelite for
self-identification to a foreigner, or by foreigners about Israelites. Its impor-
tance lies in the fact that, given the usage, ‘Abraham the Hebrew,’ biblical
critics can feel that that section of Genesis was a segment taken from some
document which had not been written by the Jews, and thus is some
independent information about the historicity of Abraham.

Now categories like ‘the Hebrew,’ where the character of the category is
that it’s used by non-members of the category and not by members of the
category except when members are identifying themselves to non-members,
are rather fundamental kinds of categories.

Parenthetically, it’s kind of a marvelous quirk about anthropology that an
enormous number of tribal names that anthropologists talk of have this
character. That is to say, the tribal names that the anthropologists write about
are very very regularly non-recognizable to the members of those tribes and
often mean things like ‘outsider,’ ‘stranger’ in some language other than the
language of the named tribe. And that’s because by and large if somebody is
told of a tribe by a member of another tribe, they get the name of the tribe
in the other tribe’s language. So, for example, Elizabeth Colson, in an
absolutely marvelous ethnography called The Makah Indians, traces such a
naming. This is from page 76.

They received the name Makah in 1855 when they made their treaty
with the United States government. The government interpreter was a
Klallum who gave the treaty makers the Klallum name for the Cape
Flaherty people. This has been the official name for them ever since.
Today most people speak of themselves as Makah though some of the
older people say they dislike the name because it doesn’t belong to them
and they don’t understand its meaning.

You’ll find the same kind of reference at page 12 of Radcliffe-Brown’s The
Andaman Islanders , and it’s a very very frequent thing.

Now we can say introductorily that there are certain categories which are

398

Part III

owned by a group other than those to whom they apply. A term like ‘the
Hebrew’ in the ancient Near East was that sort of term. Right now the term
‘Negro’ is being treated as that sort of term. That is, the attempt of the
Muslims is to treat the term ‘Negro’ as one which, if it is used, would be
recognized to be being used by a non-member. And what’s involved, roughly,
is this: We could say that what dominant groups basically own is how it is
that persons perceive reality, and that there’s an order of revolution which is
an attempt to change how it is that persons see reality. I don’t mean that in
a casual way, and I’ll try to deal with it with respect to the difference between
‘teenagers’ and ‘hotrodders.’

From time to time I mention the claim that psychoanalysis is a ‘bourgeois
discipline,’ and propose ways to see the sense of such a claim. One of the ways
one can formulate the sense of that claim turns on the way that psychoanalysis
is concerned to defend the fact that adults own reality, against, first, what they
call ‘neurotics.’ And Menninger has a magnificent phrase - he talks of
neurotics as ‘disloyal to reality.’ What neurotics are, from the psychoanalysts’
point of view, is children who can pass as adults. (And I think from the
psychoanalysts’ point of view, a psychotic is a retired child.) And what that
involves is, of course, an attempt to refuse to recognize the claims to be
dealing with reality on the part of, say, ‘children.’

Now the fact is that there are tremendous revisions of history involved in
psychoanalysis. One of the most striking is the business of the Oedipus
complex. People regularly talk about an Oedipus complex as though what
Freud says about the Oedipus plays is perfectly obvious. Whereas if you read
them without first starting with Freud’s way of looking at them, it seems to
me, at least, that it’s perfectly obvious that they’re just about the exact
opposite of what’s proposed. They’re perfectly clearly about infanticide, not
parricide. And the parricide theme is, if anything, a rationalization for an
institution of infanticide. After all, Oedipus is to be done away with by an
oracle that precedes his birth - one would presume that, if anything, there
may be some guilt about that. It’s written for adults by adults; when he finally
does away with his father he’s already an adult; they know what they’re
doing, he doesn’t know he’s doing.

There are, of course, real problems that adults have to come to terms with
about children, most of which are completely unknown, and fantastic in their
character. One is, for example, that there is, in a perfectly good sense, a
children’s culture, with its artifacts, songs, games, etc., that is unbelievably
stable. That is to say, if you look at a book called The Lore and Language of
School Children by Peter and Iona Opie, there’s good evidence that the songs
sung by London school children 400 years ago are still sung, purely orally
transmitted; transmitted without any officials taking care of them being
transmitted; transmitted where the words have passed out of the English
language. They are transmitted only by children - adults don’t even know
them. The same goes for a series of other phenomena.

There’s enormous interest to how that gets done and what it might mean
and what things like it might mean, for example, for reformulating the

Lecture 18

399

relationship of children and adults. Instead of saying, for example, that adults
are bigger and better versions of children, one might well treat them as
alumni.

Of course it’s very important that children have the kind of view they do,
e.g., a notion of dependence, because, for one, presumably if children realize
that they’re not dependent in important senses - that is to say, the norms
regulating the treatment of children are enforced by adults on adults, and
whatever children did, adults would have to take care of them because they
could be embarrassed, sanctioned, thrown in jail, etc. (and to be sure, all
adults could renounce all children, but pretty much no given adult can
renounce any given child) - they might come to see the enormous dependency
of adults on children; for example, on the response that children give to them,
like show them a little affection.

Now none of that stuff is officially recognized. It’s not recognized in just the
way that Marxists would say that the dependence of our culture on the laborer
is not recognized. It’s not only not recognized by those on top, it’s not
recognized by those on the bottom, either. And the kinds of revolutions that
then try to take place are attempts to reconstruct how it is that things are seen,
where the attempt is, in part, for some group to get to enforce a view of
themselves by others, that they administer. And a big difference between the
categories ‘teenager’ and ‘hotrodder’ is that ‘teenager’ is a category that adults
administer. What’s known about those things, ‘teenagers,’ is enforced by
adults. And of course parallel things hold for the Negro.

What I’m suggesting is that a notion of independence is trying to be
asserted against the notion of independence now being enforced. The notion
of ‘independence’ that’s held for, say, Negroes and children is, in a radical
way, asociological: They should become independent one at a time. And by
following the ways that are defined by the dominant culture. Be a good boy
and grow up, keep clean, get a good job, you can do it, it’s your problem.
That formulation of independence is enforced by the adults. Can some way be
found whereby independence can be internally administered? There are simple
and trivial solutions that people sometimes come up with, but they’re in that
genre. The attempt, say, to buy a state by the Negroes is that kind of an
attempt - they’ll define what makes for success.

What’s needed, in part, is to bring about a change in what it is anybody sees
when they see one of these things. And then to be able to control what they
know about one when they see one. Where the question of how you could
control this is, of course, an extremely complicated one. Kids know that if
they’re in cars they’re going to be seen as ‘teenage drivers.’ I’ve talked about
this phenomenon before, i.e., a kid in a car is seen as a kid in a car and not
something else, whatever he could be seen as. There are a range of categories
other than ‘teenager’ that somebody who’s a teenager could be seen as. What
are the occasions under which one or another category would be chosen
uniquely? If somebody’s in a car, and that somebody could be seen as ‘a
teenager,’ then they’re seen as ‘a teenager.’ Now what can kids do to shift that
around, or to begin to modify it? Again, what we want to have is a set of

400

Part III

modifications that are administered by members. And that means, for exam-
ple, that they will recognize whether somebody is a member of one or another
category, and what membership takes. And they can do the sanctioning.

And the character of the sanctioning does not need to involve at all that
anybody gets together and beats somebody up. But how it is otherwise
enforced is most extraordinary in a way. It used to be the case, maybe it still
is, that sportscar drivers flicked their lights when they passed each other. A
Volkswagon driver might flick his lights at a sportscar driver and be
absolutely ignored. Now it isn’t that if a sportscar driver flicked his lights at
a Volkswagon somebody would put a bomb in his car, but it was his business
to defend all other sportscar drivers on each occasion in which somebody who
wasn’t properly a member flicked his lights.

That’s much related to drag racing in the streets, because although in this
little story that the fellow tells, he pulls up in a Bonneville, another fellow
pulls up in a Bonneville and now they have a drag, while that’s conceivable,
if the hotrod culture is properly enforced, then if he does it to anybody else,
they just won’t drag with him. They will not accept him as a proper person
to drag with, no matter how fast his car is, or whatever else. It’s not a car for
doing drag racing.

Now what they do, in part, are kind of simple things. For one, everybody
by and large, including ‘teenagers,’ simply take the car they get, and drive it.
When you get the car in one or another way radically modified so that it’s not
strictly a Detroit car but something worked over by the kid, and get the
notion ‘hotrod’ applied to it (with, of course, the possibility of subversion
already present, i.e., that you get a hotrod built by some company specializing
in fixing up cars), then, for one, adults don’t drive them — or proper adults
don’t drive them - and anybody who sees one needn’t wait to see who’s
driving it, adult or teenager; they know it’s ‘a hotrod’ and there’s ‘a
hotrodder’ in it. And what is known about ‘hotrodders’ - what they do with
their cars, how they look, how they otherwise behave - are things that
hotrodders can enforce on each other to satisfy. And since hotrods are used
only by kids (or adults ‘disloyal to reality’), and once you have one what
you’re going to do with it is play with it, you become tremendously
dependent on others’ willingness to recognize that you have one, and what it
status is (’good,’ ‘the best,’ ‘rotten’). Once you make yourself a candidate,
then you subject yourself to the controls that the set of members have and
enforce.

Further, one can begin to see the interest of kids in sanctioning kids driving
ordinary cars, since those permit the retention of the ‘teenage driver’ category,
with all that’s known about it. Also, there’s a way in which some of those kids
are now doing what we don’t even have a category for. That is to say, we
don’t have a psychiatric category, like neurotic or psychotic, for something
like ‘fink.’ What it is, is somebody who’s disloyal to a lower group. I don’t
know why there’s not been a psychiatric category for ‘fink’ and ‘scab’ and
things like that; it’s kind of an interesting problem. But consider that long
utterance which follows “In that Bonneville of mine:”

Lecture 18

401

Ken : In that Bonneville of mine. I could take that thing out and if I’ve

got a tie and a — a sweater on and I look clean? ninety nine percent
of the time a guy could pull up to me in the same car, same color,
same year the whole bit, roar up his pipes and he’s in a dirty
grubby tee shirt, and the guy’ll pick the guy up in the dirty grubby
tee shirt before uh he’ll pick me up.

What’s being proposed of course is that I could put on the appearance of a
nice kid teenage driver who doesn’t drag race, and do it, and the cops
wouldn’t get me. That is, I could look like I’m loyal to them while I’m really
being loyal to you. Where of course they take it that that’s just talk, that what
he is, is being disloyal to them. 1

Now one of the things we have to see, that sets these sorts of tasks into a
framework of their real possibility, is this: We’re dealing in the first instance
with a category. They’re not ‘groups.’ Most of the categories (women, old
people, Negroes, white and the like) are not ‘groups’ in any sense that you
normally talk about groups, and yet what we have is a mass of knowledge
known about every category, and any member is seen as a representative of
each of those categories, and any person who is a case of a category is seen as
a member of the category. And what’s known about the category is known
about them, and the fate of each is bound up in the fate of the other. So one
regularly has systems of social control built up around these categories which
are internally enforced by the members, because if a member does something
like ‘rape a white women,’ ‘commit economic fraud,’ ‘race on the street,’ etc.,
then that thing will be seen as ‘what a member does,’ not what the person
with that name does. And the rest of them will have to pay for it. And
somehow these categories live with that problem, by and large. They go
around telling each other “remember, you’re a lady,’’ “remember, you’re a

1 [This note was added by the transcriber at the time the lecture was transcribed.]
Here is another story told by the same kid in another session, at which the other two
boys are not present. In this case, the proposal seems to be that he’s putting on the
appearance of a normal everyday car and its normal everyday driver while he’s really
loyal to another category perhaps similar to hotrodders, i.e., surfers.

Ken : In Mammoth my Jeep I’ve got surf stickers all over the back windows

you know?

Louise : Mm / /hm

Ken : An’ up there they hate surf. Surf is the lowe/ /st thing,

Louise : Oh I know that, heh

Ken : in the world. An’ all the adults frown upon it, the kids hate’em, they

see me, an’ they used to throw rocks / /you know?

Louise : ehheh heh!

Ken : An' I was avoiding rocks . So I finally decided this isn’t for me y’know,

I took razor blades, took all my surf stickers off? So it looked just like
a normal everyday Jeep, you know, an’ I was drivin’ around town, an’
nobody throwin’ rocks, or anything, it felt so good, y’know?

402

Part III

such and such,” and having everybody try to live up to the best image of that
thing, rather than, for example, attempting to simply change it, or ignore it,
or do whatever else.

And those systems of control are not enforced by any government. There
are no officials. Most members never know each other, but they live and die
awaiting what will be in the next day’s newspaper about what one of them
did. I would bet that you could find that in the period between the
assassination of the President and the discovery of who did it (not only ‘who’
- what name, but what he was) that every depressed group in the country sat
in mortal fear of the fact that they did it. And they knew that for those that
did do it, it would then be a fact of their lives. At best they can argue, by and
large, “Well, there are good ones and bad ones among us.”

It’s been, I think, teenagers who have made a raft of attempts to break that
kind of situation. From Beatnicks to whatever else, they’ve been engaged in
setting up independent corporations which everybody is eventually forced to
recognize on its own terms. In that regard, then, one can specifically see the
tie to drug users, who are in effect saying, “Let them all retain the perceptual
structure of reality; we’ll make our own. ” Of course, in the case of kids, the
problem of their attempted revolution is really fantastic, since they lose
members at an incredible rate. Other sorts of revolutions might have more
chance of success insofar as they can retain members.

In any event, what we want to see is that it’s not simply the case that the
fact that has long been known among anthropologists - that the Eskimos
have 17 categories of snow whereas we only have one, means that they are
much more interested in snow, or the fact that kids have 57 categories of cars
means that they’re much more interested in cars. Rather, those additions, and
the focus on them, can be ways that more or less fundamental attacks are
being launched against a culture which is stable by reference to everybody
seeing the world for what it is, without regard to whether it’s pleasant or not,
whether they come out on top or not, and not seeing that they can do
anything about it. And in that regard, the notion that it’s merely the fun of
driving fast that’s involved, which may be part of the basis for setting up
places where kids can drag safely, has provided for an enterprise which is
doomed, simply by virtue of the fact that that sets up an administration of the
phenomenon which is not theirs.

It’s in that regard, then, that the important problems of social change, I
would take it, anyway, would involve laying out such things as the sets of
categories, how they’re used, what’s known about any member, etc., and
beginning to play with shifts in the properties of a category, and shifts in the
rules for use.

Lecture 18

403

This lecture was given on March 25, 1966. Several months thereafter, Sacks
clipped an article from the New York Times onto his copy of the transcript, with a
small segment bracketed.

New York Times , Monday, August 8, 1966

Harlem Residents Seek to Dispel Whites’ Fear of Area (continued from first
page, second part)

“Ten years ago, a Negro didn’t want any part of black,” a young man said.
"He wanted to be as much like the white man as he could be, but now he’s
realizing that he never will be some kind of white man. We’re finding an
" identity that has eluded us for so many years - as black men.”

L A surprising number of people in Harlem now use the word black, in
preference to Negro, in speaking of their race. Some are turning to African
history and African dress to give themselves a more buoyant frame of reference
than the melancholy history of black suppression in America.

Lecture 19
Appearance verbs

I’m going in part tc deal with “and if I’ve got a tie and a sweater on and I
look clean.’’ I want first to consider the sorts of things that can be involved in
“I look clean.’’ Its interest lies, in part, in the fact that it’s partially related to
a class of apparent paradoxes that there has been a good deal of philosphical
literature about. Also, it’s interesting in relation to those verbs that name
activities, which we talked of using Austin’s term, ‘performatives,’ e.g., “I
promise” and the like.

One of the things we noted, and that’s important about the activity
involved in using performatives, is the marked differences between the
activities involved, given differences in the person (I, he, you) attached to the
verb. “I promise” is one sort of thing, a way of doing promises. “He
promises” and “you promise” are quite different things, and in part the
warrant for saying “he promises” and “you promise” turns on the person
being referred to having said something like “I promise.” So the first person
could be doing an activity, whereas a second and third are reports on it, which
are warranted by there having been a first-person assertion (and that could
have been other things than saying “I promise,” which is not the only way to
promise).

Now there are a class of verbs which we can call ‘ appearance verbs’ that
work in a kind of reverse fashion from performatives. There is again a split
between T and ‘you,’ ‘he,’ ‘she,’ but it’s a different sort of split. And perhaps
not seeing the split, and not seeing the way uses of T get warranted with the
appearance verbs, is a basis for the sorts of apparent paradoxes that have been
proposed; things like “I am dreaming.” Another is “I talk in my sleep.”
Things that turn on sleep are very good ones and very usual ones for
consideration of how is it that somebody can say that, and if they say that
don’t we have a paradox of sorts, where if you were dreaming you wouldn’t
be in a position to say “I am dreaming,” etc.

So, recalling, to start out with, that a warrant for the use of ‘you’ or ‘he’
plus performative would be somebody’s use of ‘I’ plus a performative, let’s
then consider these appearance verbs, where what I’m going to propose, in
arguing that there’s a kind of reversal situation going on, is that a warrant for
the use of ‘I’ plus an appearance verb is the occurrence of somebody’s use of
‘you’ or ‘he’ plus the appearance verb. Simplest case is, a warrant for saying
“I talk in my sleep” is that somebody said to you, “You talk in your sleep.”
If one can move from ‘I’ plus appearance verb to get the derivation of it in
a ‘you’ plus appearance verb, then of course there’s no paradox.

The paradox turns on the fact that one has a normal class of T plus

404

Lecture 19

405

something, where it’s warranted by the fact that I know it, or I’m the best
authority, etc. And one doesn’t separate out those which are normally or
properly arrived at by somebody telling you it. And the grounds, of course,
for not seeing that it has that derivation are, in part, certain sorts of
paraphrases that are permissable. Now, the phenomenon of paraphrasing is a
fantastically important one. It’s learned very very early. In Children Tell
Stories there’s at least one use of it in an under-three-year-old’s story. The kid
says, “His mommy came home and told him to get out of the street.” That
statement is recognizable as not merely a quotation, but a paraphrased
quotation. Mommy wouldn’t say “to get out of the street.” And it’s of course
a kind of interesting thing to see how kids would ever learn to make that sort
of paraphrased quotation.

The phenomenon of quotation in general is an extremely tricky problem,
especially in languages like English where there are no special ways of
invariably marking that what we’re doing is a quotation. In some languages
there are grammatical markers that say ‘a quotation is coming’ or ‘this is a
quotation.’ In English what you can do is say “He said” and the quote
follows. But you don’t have to, and there are considerable number of activities
which turn on ambiguously using quotation marks. For one, it’s a way of
being ironic. For example, how would we decide that in the statement “You
must be crazy or you wouldn’t be here,” he’s asserting ‘You must be crazy’
or he’s asserting ‘You must be “crazy”.’ That problem has been one which
has had rather vexing kinds of consequences in various historical disciplines.
I’ll deal with it when I consider some problems in the data, of deciding that
we have a quotation. I only mention it now.

Now, given that T can be used directly, it may be tricky to see that it’s
being used as a paraphrased quotation, i.e., that if one says “I talk in my
sleep,” what’s being said is something like: “John said, ‘You talk in your
sleep’.” “John said I talk in my sleep.” “I talk in my sleep.” Where, then,
the first-person usage is a paraphrased quotation, and the warrant for the
assertion in the first person is some assertion in second or third person. Of
course this doesn’t hold for all appearance verbs. But for those in which what
one has is a usage which has the warrants I’ve laid out, there is clearly nothing
paradoxical involved. And it is, of course, to be noted that Members perfectly
well know this possibility, and at least on some occasions recognize a case of
first-person usage to be a quotation. So, in the group therapy data, in another
session one of the kids is saying “I’m crazy” and the therapist says, “You keep
saying you’re crazy. Have people been telling you that?” Another: A little girl
comes home and says to her mother, “I’m beautiful.” Her mother might well
say, “Who told you that?”

We can then mark as the class of ‘appearance verbs,’ those for which it’s
appropriate upon the assertion in the first person for someone to ask “Who
told you that,” and where it’s the case that a warranted use is that someone
told you it. For any of those, then, there is in principle no paradoxical
character to them. And the kinds of troubles that can occur are simply things
like saying it without the proper warrant (i.e., without someone having told

406

Part 111

you), or treating other people’s assertions at face value when they’re just
trying to get things out of you or they’re just trying to degrade you, or taking
it that they think so though they haven’t said so, etc.

We also can see that we get another pronoun relationship, this use of T
warranted by some use of ‘you’ or ‘he’ or ‘she,’ where again we have a kind
of screwing up of the pairings that we started off with as obvious, i.e., I-We,
you-you, he-they, etc. We now have T derived from ‘you’ or ‘he.’

And of course we have a kind of obverse to that classical philosophical
problem, How do you know others’ minds? which is. How do you know your
own appearance? And the problem, How do you know your own appearance?
turns on the fact that statements about your own apperance are warranted by
virtue of somebody else’s assertion. Where those assertions are problematic,
not merely by virtue of their correctness or not, but by virtue of the set of
affiliated activities that can be being done when somebody says, for example,
“You look nice.’’ And since such assertions are a basic way of learning about
one’s own appearance, a consequence of, e.g., ‘taking all praise with a grain
of salt’ is that you get a rather special picture of yourself, since each of the
decent characteristics can be heard as ‘mere praise,’ and the like.

Now in the case of “I look clean,” ‘clean’ is not clearly one of those, since
it can be learned, and there are devices for seeing and saying about yourself
that you ‘look clean,’ on your own behalf, or by use of the mirror. The fact
that some of the appearance verbs can be decided by oneself and some can’t
may well pose classification problems, and pose trouble for Members’
knowing which ones they can decide and which ones they can’t. And the
metaphor of a mirror gets, of course, enormously troublesome. Psychiatrically
you get a large variety of misunderstandings of proper classifications, which
are symptoms. That is, from treating none of the appearance verbs as things
you have rights to decide about yourself, to perhaps allowing nobody rights
to decide for any of them, i.e., assimilating them all as things like ‘clean.’
From ‘I can’t trust myself at all’ to ‘I can’t trust anybody else at all.’

Let’s go back now for a bit. We have an enumeration of features: “I’ve got
a tie and a sweater on and I look clean.” What controls the enumeration?
There are, of course, many places where we get enumeration of features. I take
it that in this case it’s unanalytically apparent that the enumeration is
controlled by a picture being offered which is directed to presenting an
appearance type, i.e., “If I look like a nice kid” - a ‘nice kid’ in contrast to,
e.g., a ‘hotrodder.’

The ‘nice kid’ category is not generically stable in that, that for which some
persons - adults, cops - would see ‘a nice kid,’ may well provide for
hotrodders that they’re seeing a ‘Soshe,’ a ‘square,’ whatever else. And one
other thing which we will see, which is related to the appearance of the ‘nice
kid’ for the cops or adults in a deep way, is that he doesn’t want to be seen
only as ‘a nice kid,’ but he wants to permit being seen as ‘a nice kid’ by the
cops, and an eligible dragster by kids. What he wants, in part, I take it, is to
insist on the difference between the appearance of the ‘nice kid’ for cops, and
the appearance that can be seen by a hotrodder, who should not be ‘misled’

Lecture 19

407

by his apperance. That ambiguously intended picture could be permitted by
a legitimate dragger. It could be refused, by seeing him as in fact a nice kid
and not just one who wants to appear so to 'others,’ i.e., in fact ‘a Soshe.’ Or
it can be seen as an ambivalence on his part. In this case, perhaps, that’s what
the term poor little rich kid’ intends, i.e., somebody who wants to make it
both ways, with adults and with the kids.

And you’ll notice how in the story here, this little hypothetical anecdote, he
gets it both ways. And he sets it up in a nice way. And that is that he’s driving
along and he doesn’t make the move to do a drag, but somebody else does
it. Where of course somebody else doing it involves somebody seeing him as
acceptable, where that somebody else is, if not acceptable to an unquestion-
able hotrodder, at least more acceptable than him, i.e., the somebody else is
wearing “a dirty grubby T-shirt.” Where of course the offering of a signal -
roaring up his pipes, which is the signal for a drag - involves the person who
does it proposing himself as a candidate dragger for someone who he takes to
be another - and not merely another candidate, but one who is entitled to
accept the candidacy of the first. He might have said ‘‘I pull up to him, roar
up my pipes and he accepts,” to which they could then say "Yeah, you might
do that, but nobody would accept.” So it may well be not incidental that he
sets it up with the other fellow making the offer. The other fellow would have
had to decide that he was one who would be willing or competent.

Q: Are people aware, actually aware that they do this in conversation, such
as having the other person offer the invitation by roaring up his pipes? Is it
your opinion that he knows what he’s doing or is it something that they have
learned internally in conversation, and never become aware of it.

HS\ Well, what I understand by knows’ is that he does some such things
as require my analysis to put that into him. See, you have a notion of ‘a brain’
and ‘a person’ and ‘what they know.’ And your issue is, ‘Well, do they really
know this or that? Am I going to get my knowledge corrected here? I know
that’s so, I know this is not so.’ As I said when I began out, and as I intend
throughout, the task is to find what it is that has to be built into somebody
to make them someone who can do these things. That what they know, they
know, quote, consciously or whatever else, is a problem. I don’t at this point
intend to solve it. I don’t know how to solve it. I’m not looking at it. The
question is, what kind of a machinery do I have to put in there? Now, a
machinery that would say about something that, not only is it there, but it’s
there as a matter of their awareness, is another piece of apparatus, perhaps.
But I have no idea. I don’t have any idea of what it is that being conscious
of something would affect. I suppose it’s a kind of heritage from Socrates that
one ought to be conscious of all these things, otherwise you can’t really control
them, or something like that - which supposes that if you are conscious of
them you can control them.

Okay. So at least for our story’s purpose, what he does is partially construct
an appearance which is still satisfactory as an acceptable partner in doing drag
racing. It’s also an appearance that he takes it works for the cops, since the cop
also observes the scene and doesn’t arrest him. Now it’s kind of nice to watch

408

Part III

how, given the fact that the whole thing and all of its parts are hypothetical,
he sets it up in such a way as to show which of the hypothetical parts are
criterial of the outcome. That is, each clause could be prefaced with ‘if’ just
as legitimately as any other. Again, the data:

Ken : In that Bonneville of mine. I could take that thing out and if I’ve

got a tie and a - a sweater on and I look clean? ninety nine percent
of the time a guy could pull up to me in the same car, same color,
same year the whole bit, roar up his pipes and he’s in a dirty
grubby tee shirt, and the guy’ll pick the guy up in the dirty grubby
tee shirt before uh he’ll pick me up.

So, e.g., “I could take that thing out” could be rephrased as ‘If I were to . . .’
The only place where he does use ‘if is to mark this appearance sentence, “if
I’ve got a tie and a - a sweater on and I look clean,” where the appearance
difference is parallel to the appearance of the other, and that’s supposed to be
a criterial matter, as to how one ever gets arrested. So the ‘if seems to signal
that thing which is the hypothesis to be tested by whatever happens. And it’s
nicely set up as a hypothesis - all other things are constant, even though
they’re all equally hypothetical.

One thing you might note about this very long sentence is, apparently if
you build sentences out of phrases and clauses you get a lot more positional
freedom into them than when you don’t use those sorts of things. What I
mean to note is, e.g., here’s this phrase, “99 percent of the time.” It doesn’t
modify what follows it. It’s not “99 percent of the time a guy could pull up
to me . . . etc.,” but, when we put the sentence back together again after
having heard it, “if I look clean, a guy could come up to me in the same car
. . . etc . . . and 99 percent of the time he’ll get picked up.” It’s kind of
interesting that we can really rearrange those matters; whereas within the
clauses the words may not have that kind of positional freedom.

Now, what about ‘look’? I suggested that what the enumeration is directed
to do, is to offer the appearance of ‘a nice kid,’ where what’s very important
is that it’s to be seen as the appearance of a nice kid who is not a nice kid. Not
a nice kid, that is, for possible hotrodding, because a nice kid wouldn’t do it.
Now I have the feeling that the word ‘look’ there, which, to be sure, is
legitimately usable with ‘clean,’ stands as an alternative to ‘am.’ In the case
of clean, it’s generally quite weak to say “I look clean,” since if you look clean
you most likely are clean, and the difference isn’t very important. But in the
case at hand, being it and looking like it are importantly different; “I am
clean” could then have said of it, “Yeah, you’re a nice kid,” “You’re a
Soshe.” Sticking in the ‘look’ there may well mark the fact that he’s not really
the thing that he’s reporting the appearance of, but only looks like it.

Now there are certain questions which are kind of relevant to the theory he
seems to be using here. For one, if it’s the case that he’s possibly doing a drag
with somebody else, and a cop sees it, then for the cop, the decision that he
is doing a drag is, in part, decided by whether he looks like somebody that

Lecture 19

409

would do that. It’s not the case, apparently, that if he does something that
looks like it, if he’s doing something that looks like what the other person is
surely doing, then he’s doing it. And it is, of course, regularly the case that
when some activity is possibly going on, Members decide between whether it
is or isn’t going on by virtue of some determination as to whether its
participants are ‘the sort who do it.’

The use of such procedures are quite relevant, for example, to get statistics
which show higher crime rates among Negroes, or among lower-class kids. If
such a procedure is involved, then results are provided for by the way it is that
one looks to find cases. That thing done by ‘hotrodder’ can be seen as
‘dragging,’ if it might or might not be. If he doesn’t seem to be ‘a hotrodder’
then it might not be ‘dragging.’ In which case, you find hotrodders and nice
kids having differential arrest rates. And it’s not simply that that’s available
to us, it’s apparently available to them.

Lecture 20

Character appears on cue ; Good
grounds for an action; An
explanation is tjoe explanation

I want to focus for awhile on . . roar up his pipes and the guy’ll pick the
guy up ...” In the first instance, there’s no problem about “the guy’ll pick
the guy up,” as to who the guy who picks the guy up is, by virtue of ‘picking
somebody up’ being category bound to ‘cop,’ permitting, then, a replacement
of ‘the guy’ - which would be a general pro-term for ‘male’ or whatever - by
‘the cop.’

When I was considering “The baby cried. The mommy picked it up,” one
of the things I was remarking there was that when a character who has some
proper grounds for appearing and some proper thing to do, has its cue, then
there’s no need to account for how they happened to have come on the scene.
Given “the baby cried,” you don’t have to get a characterization of where the
mommy was, that she was anywhere around; the mommy does her job. 1 And
we can notice that something akin to that 'character appears on cue
phenomenon is present here. Where the cop was, how he came on the scene,
is not indicated and not missing.

In certain circumstances there is a variant of the ‘character appears on cue’
rule, which is much stronger than the above. And that’s what I want to focus
on for a while. We can contrast what I want to develop with the phenomenon
of “The baby cried. The mommy picked it up” by noticing that not only are
there a large amount of other circumstances under which a mommy might
pick her baby up besides the baby crying, which are also ‘on cue,’ but there
are also plenty of circumstances where a mommy could pick her baby up
without having any such cue from the baby. Mommies are free to pick their
babies up, perhaps at will.

Now, for certain characters and certain activities of those characters, we
have a circumstance where it’s not merely that they can appear on cue, but
they properly appear and do their activities only on cue. And that fact has a
series of very large consequences. First of all, it permits a Member’s
description to be adequate which reports the occurrence of the character
appearing, and his activity, without having to indicate what the cue was,

1 See Spring 1966, lecture 2, p. 254. The foregoing part of this paragraph was
taken from Fall 1965, Lecture 9, p. 183, as it is a bit richer than the consideration
in the original lecture 20.

410

Lecture 20

411

where it will be heard that there was a cue. So, in the case at hand it isn’t that
the guy roars up his pipes and that’s why he gets arrested, but it is by virtue
of roaring up their pipes being a signal to drag that, presumably, then, the
drag would take place, that generating a violation - speeding - and the cop
then picking him up for that.

In a way, that consequence is secondary to the strong character of this
variant of the ‘character appears on cue’ rule. I’ll quote a piece of data and
we’ll consider the strong inferences that are permitted if a character who
properly appears only on cue and does his activities, is said to appear. This is
a conversation between a psychiatrist and a person who’s calling for help. It’s
a first conversation, perhaps the fifth interchange into it - no more than that.
‘A’ is the psychiatrist. ‘B’ has had a quarrel with his wife, and now he’s
reporting a scene that took place.

A: Yeah, then what happened?

B: Ok, in the meantime she ((she is the wife of B)) says don’t ask the

child nothing. Well, she stepped between me and the child and I got
up to walk out the door. When she stepped between me and the child
I went to move her out of the way. And then about that time, her
sister had called the police. I don’t know how she - what she-

A: Didn’t you smack her one?

B: No.

A: You’re not telling me the story, Mr B.

B: Well you see when you say smack you mean hit.

A: Yeah you shoved her, is that it?

B: Yeah I shoved her.

The power of the variant of the ‘character appears on cue’ rule for this
special class of characters and their activities is that one who does not know
the person they’re talking to, beyond a very brief acquaintance, can tell that
the other is lying. That is, when the other proposes both the occurrence of the
character’s appearing, and no cue. And that is to say, if the characters appear,
then there was a proper cue. We can see, then, where this strong rule holds
by seeing those places where a hearer can say that the reporter is lying.

Now that’s quite an extraordinary kind of fact. After all, the teller of this
story is the one that presumably knows what happened. And you don’t have
to know that he’s a known liar. That he proposes such a character appeared
and there was no cue, is good enough grounds for asserting to him that he
lied. And that assertion can work, furthermore. That is to say, he can
recognize that you have good grounds for knowing that he lied. He doesn’t
say, “How in the world would you know?”

It is important to see that, again, it’s ‘recognition’ that we’re dealing with
- here, recognition of a lie. And it’s perfectly conceivable that one could offer
as the cue, something which wasn’t true, either. So a successful lie under these
circumstances would be to propose that some adequate cue, other than the
real one in this case, occurred. Where, then, what we want to see is how it is

412

Part III

that the phenomenon of a lie becomes not idiosyncratic as to its recognition.

In part, the importance of the phenomenon can be perfectly obvious by just
considering data like this. But it has provided a focus for rather more
extended concern. The basic sorts of problems that seem to be focused around
it can, at least roughly, be readily enough seen by considering one funda-
mental myth which turns on this phenomenon. And that is the story of Job.

Recalling the story of Job, one of his problems was essentially as follows:
He’s a good man, and he’s well rewarded for being a good man. Unbe-
knownst to himself, he’s being tested. The test consists, in part, of a
destruction of the possible bases for his being a good man, i.e., the rewards
- his wealth, his family, etc. And the occurrence of that destruction poses for
Job a problem: Why did it occur? That it is a test is unavailable to him. He
is quite convinced that it isn’t a punishment; convinced because he doesn’t
figure that he’s done anything to warrant that.

Now come his ‘friends,’ and they’re engaged in a series of dialogues, which
are, in part, directed to the relationship between morality and rewards in this
world. They take it that what it is that would have warranted what had
happened to him is that he did something punishable - where what has
happened to him is properly to be conceived of as ‘punishment’, given that
unless what has happened to him can be conceived of as ‘punishment’, it’s
inexplicable. It has, therefore, to have been what it might have been, so what
he ought to do is confess. If he admits his guilt, they propose, he’ll be in good
shape.

We have, then, this phenomenon: If an explanation is available, then it’s
that explanation that is the explanation, and formulates what it is that’s
happened. Now, it’s a very general phenomenon. Let me quote some further
data which lays that aspect of things out quite clearly. It’s a decision from a
labor arbitration case. I choose it only because it’s very explicit in its
formulation.

The story of the other discharged employees approaches the bizarre.
Twelve of them testified before me. Each of them claims to be a
completely innocent bystander, wholely at a loss to understand why he
was picked up for discharge. None of them admits being part of the
crowd in any of the demonstrations. None of them admits even the
normal curiosity of an innocent bystander. Each claims that when the
lights went out or the line stopped, he asked his foreman what to do,
and upon being told to stay or go home as he pleased, but that his time
stopped in any event, he left for home. One of the men, a lively young
boxer, asserts that after seeing the crowd and the excitement, he calmly
repaired to a warm spot and went to sleep. He did this, he asserts, on
two of the three days, being absent on the third, and slept the peaceful
sleep of the just, until the excitement completely quieted down. All, it
seems, were veritable angels, above and beyond contagion by the
excitement in the department.

Lecture 20

413

Now there unquestionably were serious stoppages in Department 84
on November 5th, 6th, and 8th. There were vociferous and angry men
milling around and demanding action. Who were the incensed men
who did take part? Who were the angry men whom it was difficult to
get back to work and who were so incensed that, as the Union claims,
they turned against their own committeemen and even assaulted two of
them. How indeed were these 14 chosen?

The Union advanced no explanation. There is no suggestion that
these men were chosen by lot. And even such a method would normally
be expected to catch some of the guilty. And there is no basis whatever
in the evidence to suppose that the men were selected because of any
personal animosity against them, with the slight possible exception of
one man. Nor were the men generally regarded as troublemakers of
whom the company would be glad to be rid. The company’s explana-
tion is simple, and without any contradiction other than the incredible
stories related by the men themselves. The labor relations conciliator,
with the help of his assistants, took the names or badge numbers of the
most active men in the crowds that demonstrated in his office. This
accounts for twelve of the fourteen. Under these circumstances I cannot
give credence to the men’s protestations of innocence.

Now the notion that it would be the business of a person proposed to be
guilty, to explain how he came to be selected as the one proposed to be guilty,
is crucial to these formulations - crucial to Job, crucial to the detection of a
lie, and, of course, part of the enormous power of this variant of the ‘character
appears on cue’ rule. We have, then, a circumstance where it is not merely the
case that, lacking reference to the cue, some cue can be assumed to have been
present, but the assertion of ‘no cue’ is an extremely difficult business to
defend - and it’s the business of the asserter to try to defend it if he can.

The phenomenon has further relevance for the data that we have. It is that
sort of thing which provides for the occurrence of the question, “you must be
crazy or you wouldn’t be here.” And it makes for a series of extremely difficult
problems for any given person to handle, e.g., once you’re in a mental
hospital or a prison, though it may be the case that each Member takes it that
they are there wrongly, they also take it that everybody else is there rightly.

Indeed, the very fact of a hearing is extremely troublesome to deal with.
The notion that, for example, if possible mental patients were assured
hearings beforehand, then that would remove a great deal of injustice, while
it may be so that it would remove some injustice, has, if it’s going to be
argued, to take account of the fact that the mere presentation of them as
possible ill persons, possible criminals, is a rather different thing than saying
that anybody is possibly ill. ‘Anybody’ isn’t there, and there must have been
some reason that this one is there. They need, then, not simply provide that
they’re perfectly okay, but explain how they got selected - where that
explanation is extremely tricky. Because, for one, it has to allow that such
procedures as selected them will, by and large, in almost every other case,

414

Part III

select people properly. Were they to argue otherwise, i.e., that every case -
theirs being only an instance - is selected in God only knows what way, then
of course they prove their insanity.

And this is without anybody knowing by having studied how it is that in
fact, in any case, persons are selected. The matter of reforms in these areas is
by no means as simple as is sometimes suggested. Because, for one, a reform
that would involve ‘a really true innocence until proven guilty’ - in the sense
of no supposition of guilt by virtue of having been selected - would serve to
make activities like those of the police extremely difficult for them to do. And
that is even if such a notion could be maintained, and I don’t see how it could,
anyway.

Again, what’s involved, in part, is the circumstance under which various
persons other than those who did some observation, have the seer’s maxims
available to ‘see’ what it is that must have been seen by those who did the
observations. Recall when I discussed “The baby cried. The mommy picked
it up,” I simply asserted that producers of activities could use seer’s and
hearer’s maxims. That the seer’s maxims are available to producers of actions,
and are crucial to them in producing their activities, can at least in part be
gotten at by locating those things which Members have available to them as
things that can be done subversively — under the very general usage of
‘subversion’ I offered. And in that regard, then, we get, for example, “If I’ve
got a tie and a sweater on and I look clean,” where what’s being proposed is:
‘I could put on an appearance for which I can be assured as to how it is that
I will be seen by somebody unknown to me personally, via the fact that he’s
a cop, and via the fact that he will use some seer’s maxim to see who I am,
where I can use those facts to generate an appearance which has some more
or less guaranteed reading.’

Where ever, then, Members know how to do subversive activities, we can
take it that they have available to them some notion of what seer’s maxims
will be used to observe them as problematic phenomena. Where, of course,
their subversive activities are directed to satisfying some constraints on how it
is that persons are to be seen. In that regard, then, subversion is simply a
consequence of seer’s maxims, hearer’s maxims and the like. If you want a
grandiose version of it, this is so in the sense that a poet is subversive according
to Plato, since he knows how to produce something that sounds real while not
being real. Which is to say, he knows how to produce something which will
be heard as correct though it isn’t. Plato claims they’re liars because he takes
it that they’re not seen as lying. And successful lies satisfy constraints for
producing something that will be heard, according to one of those maxims,
as ‘correct description.’

Let’s move on a little bit now. The next line is “But not many people get
picked up in a Pontiac station wagon.” Notice that there is transformation in
the name of the object being referred to. Earlier ‘Bonneville,’ now ‘Pontiac
station wagon.’ It’s the same car. The notion that names are used only as
conventions seems unuseful here. Something like a notion that Russell offers
- that names are disguised descriptions - would seem to be much more

Lecture 20 4 1 5

appropriate. That names are doing things, and that the use of a different name
than another is criterial, seems clearly to be the case.

There’s a beautiful example earlier in the session, exactly parallel to this
one. The fellow that’s talking about his ‘Bonneville’ is there talking about his
Jeep. (In fact, the car that’s his in the family, is a Jeep. The Bonneville is his
mother’s.) And he’s saying that he just put a ‘Thunderbird’ engine in his Jeep,
at which point somebody else says, “Why the hell do you want to put a Ford
in a Jeep?” So, his ‘Thunderbird’ becomes a ‘Ford,’ as his ‘Bonneville’
becomes a Pontiac station wagon.’

Focussing on the statement “But not many people get picked up in a
Pontiac station wagon,” let me first note that if you took that sentence
isolatedly, you could come up with a whole series of possible hearings that
would never occur to you dealing with it as we are. For example, I take it it’s
perfectly obvious that we do not hear the statement as asserting that the
Pontiac station wagon is not one of those circus cars with lots of people in it,
i.e., that when a Pontiac station wagon gets picked up it’s relatively empty,
though syntactically that’s a perfectly decent reading of “not many people get
picked up in a Pontiac station wagon.” And if you’re dealing with possible
syntactic ambiguities, you would have to come to terms with how it is that
that would just not occur to you. And the notion of ‘the sentence’ as subjected
to syntactical consideration as a sentence isolated, and that giving you
important results, has to be considered when you find how strange, indeed,
such a reading would be - though of course there might be some use of that
sentence which would have such an intention.

It’s clear, then, that we have to deal with the sentence in its position in this
argument, in trying to see what it is that’s being done with it. Where one of
the facts is that what’s being done with it is perfectly clear to the hearer of it,
who says right off, “Now well I agree it looks like a Daddy - it looks like a
damn Mommy’s car.”

Now there’s a point to the statement somewhere, and the question of how
to find how that statement does that point is a very large issue. I can’t deal
with it. But there are some things, anyway, that might be said.

It stands in nice juxtaposition to the earlier pair of statements, where the
fact that you’re bound to get picked up in a hotrod is treated as posing the
challenge. Here, “But not many people get picked up” seems to involve, in
part, that if you don’t get picked up, then that’s not countable as a success.
The conditions under which not getting picked up is countable as a success,
turn on those conditions which provide that you’ve got a challenge.

Once we get to that, what seems, then, to be involved is that we have here
- as we have throughout - a piece of an argument serving quite adequately
to suggest clearly what the argument is. In this case, for one, what’s crucial
is not simply that not many people get picked up in a Pontiac station wagon,
but the relativity of that fact to the drivers of Pontiac station wagons. That is,
plenty of people drive Pontiac station wagons; of them, few people get picked
up. Where, if few people drove Pontiac station wagons and few got picked
up, you could still have a perfectly good challenge, insofar as Pontiac station

416

Part III

wagons were treated as relevant to whether you get picked up. And that is
part of the apparent argument, as seen by the respondent to it. Pontiac station
wagon is at least a case of a type: Daddy’s car, Mommy’s car. Where users of
that type are treated as having a fate in common, such that certain accounts
and only certain accounts are relevant in explicating certain problems. For
example, why you didn’t get picked up is to be handled via the fact that you
were driving one of those cars, and other possible explanations are ruled out.

And again, that argument that the fate turns on the type of car is one which
involves the fact that seer’s maxims are available even for this hypothetical
case. Some such thing as: ‘A policeman sees a Daddy’s car or Mommy’s car,
and knowing what he knows about such cars and how they’re driven, and
knowing that they’re a legitimate type for him to use to decide what they
might be doing, ignores them.’ So that a person in one of them who violates
a law has a much lesser chance of getting picked up than a person in some car
known as a type driven by systematically violative drivers.

This knowledge being available means that when we deal with arrest
records of teenagers, we’re not dealing with such things as might be handled
via Driver’s Education and things like that. They know their fate before it
happens, and indeed they’re proposing to reject, not merely the chance not to
drive fast, but the chance to drive fast under circumstances in which they’re
not so observable. That is to say, the chance of getting arrested is apparently
some strong consideration in doing whatever they do, in the ways that they
do it.

Lecture 2 1

Misidentification; Membership
categories ; Utterance pairs ;
Paradoxes

In this lecture I’ll be concerned primarily with two technical problems. The
first is as follows. Is it the case that for some places, the sequence in which
some rules are applied to give us an analysis is important? And what, then, are
the proper sequences for applying the rules to give us an activity? And the
second concerns the orderliness of relationships between more or less complex
activities. Can we find order in their relationship and can we provide for that
order?

By ‘complex activities’ I mean things like doing discussing, which doesn’t
take place in an utterance but involves at least several. What I’ll be concerned
with is that in the given segment that we’ve been considering, there are two
complex activities which are named as taking place, by the participants.
‘Discussing’ is one, and the second is ‘tearing down.’ One question we want
to ask is whether such activities as those occur recurrently in the order they
occur in here, and is there some way of showing why they occur in that order,
and how one leads to the other? We’ll get to that problem in a while. I’ll deal
with other things as well, besides these two problems.

We can begin to consider the problem of sequences in applying rules by
looking first at this line, “Yes Mommy. Thank you.” Recall that I’ve
mentioned the problem of what could be called ‘ misidentification ;’ that is to
say, that a term like ‘Mommy’ is used to address and refer to somebody who
isn’t the mommy of the person so referred to. And when I’ve mentioned that
problem I’ve said that it has some considerable theoretical importance. One
importance it has is something like this: If we took it that the way to
determine who it is that’s being referred to by some category is to be decided
by a ‘ correspondence criteria ,’ i.e., by looking to see who it is that’s properly a
member of that category, then the problem of misidentification would be
extremely troublesome. We might end up simply with saying that either the
person who could be correctly referred to by that category is being talked to
or of, or, if there seems to be nobody who could be so referred to, then an error
is being made. And we might not have a good way of dealing with intentional
misidentification , which is a way of doing some particular activity, e.g.,
‘insulting.’

This is a condensation of lectures 21 and 22.

417

418

Part III

If a correspondence criteria has this sort of inadequacy, that it cannot
deal with intentional misidentifications, then we have another way - besides
the way I formulated in ‘The Search for Help’ paper - of coming to see the
systematic, general, in-principle inadequacy of the social scientists’ use of
correspondence criteria for deciding who it is that could be being talked of
by the assignment of some category that could be correspondence-correct.
That is, another way besides the way I posed there, which turned on two or
more devices being available to categorize anyone who might be being
talked of.

When we consider the correspondence criteria, and pose for ourselves the
problem of assigning a sequence to a set of rules for reproducing the activities
being done, we might use the correspondence criteria as one possible
first-order rule for dealing with the use of membership categories, that would
run something like the following: ‘If some person uses a membership
category, then look to see whether there is somebody present, properly
characterized by that category. If so, treat them as the person categorized.’
And there are clearly places in which such a formulation would work.
Suppose a child is in a room with his parents and says, “Get me a drink of
water, Mommy.’’ Then one could take the term ‘Mommy,’ see who it is
among the other persons present that could be categorized by that, and
arrive at one. And there would be a variety of ways of arriving at that one.
Then, of course, treat that one as the one being talked to. Such a rule
wouldn’t work to give us a person being categorized in this case, and it
wouldn’t, I take it, deal with terms which are different from ‘Mommy,’ like
‘stupid,’ which could be used to refer to and address somebody in some
utterance.

Now the question is, does there seem to be some way that Members go
about locating who it is that’s being addressed by some term? As I pointed out
earlier, and this is of course perfectly apparent - though that it’s apparent
does not mean it’s unimportant - who it is that’s being addressed by “Yes,
Mommy” is understood by the Members, and they then may well seem to
have some way, apart from correspondence correctness, to determine that
someone is being referred to by a membership category. How, then, do they
do it?

I think that what’s relevant for a solution to the problem is that the rules
of conversational sequences are the first rules to be used. Then what one wants
to see is whether there are some rules of conversational sequencing which,
before considering the membership that was used, will tell you who it is that’s
being addressed by that term. Then the term that was used can be considered
as to what activity it might be doing, given that you know who it is that’s
being talked to or of, with it.

So what we need, then, are the variety of appropriate rules of conversation
sequencing. And these can be more or less abstract, or they can be particular
to the sorts of things being done. Where their abstractness can just provide for
a testing of the rules.

Now, we’re at this point where we have:

Lecture 21

419

Roger : Ken, face it. You’re a poor little rich kid.

Ken : Yes Mommy. Thank you.

And for “Ken, face it...’’ we can, for example, take it that one rule operating
is: A person addressed by some utterance is given the occasion to talk next.
Now that’s very loose and general, and doesn’t tell us what in particular is
being done by the thing they’re addressed with, or what they might be doing
next, or, for that matter, whom they might talk to, if there’s anybody in
particular that they might talk to. But it is recurrently the case that that is
done, i.e., giving a person the floor by addressing them - particularly by
addressing them with their name. (You can use it to get the floor yourself, in
a variant of that situation I considered for kids, i.e., “You know what,
Daddy?” gives the floor to Daddy who gives it back. So in this particular
two-hour session we regularly find things like: Somebody says, “Roger,”
Roger says “Yeah,” and the other talks.)

As it happens in conversations in American English, it’s regularly the case
that the use of a name in address, signals a change in the topic of conversation.
I don’t know why that’s so, but empirically it looks to me like that’s done
very regularly, and it seems to do that. There isn’t any grammatical way to
signal a change of topic in English, though there is in other languages. Some
American Indians apparently have a grammatical means of signalling that
you’re going to make a topic change. In this case I take it that we could find
that a topic change is being introduced with this utterance, from an
‘automobile discussion’ to the beginning of ‘tearing down,’ where Ken is the
topic.

So: Ken is addressed, Ken can speak. Secondly, at least one of the things
that seems to be done with that line “you’re a poor little rich kid” is to offer
an insult. Insults come in pairs. To a first, a second may be returned. We can
get at that in a fairly decent way, in that, for one, even if we’re uncertain as
to the fact that the first was an ‘insult,’ there are a collection of ‘second
insults,’ i.e., locatable returns to insults. And misidentification of kinship
terms are, very regularly, just such things. That’s a very very ancient
convention. Here’s a citation. Eric Partridge, in The Shaggy Dog Story (1954),
page 15, reports as an ancient Greek story (and the same thing has been
reported very widely):

A pert youth, meeting an old woman driving a herd called, “Good

morning, mother of asses.” Return: “Good morning, my son.”

There are, of course, fairly decent formal reasons for using kinship terms in
this fashion. If it is that something derogatory is being proposed about a
person, then one way of generalizing it, or providing for its source, is to use
kinship terms.

Now if we have any first utterance - an insult or something else - that
provides for a proper second, and provides also for whom that second should
be addressed to, then what we can do is see, first, whether it looks like the

420

Part III

rules are going to be used by virtue of the fact that the person located by the
first speaks. If they do, we can look to see whether or not they include a
membership term, and if they do, we can take it that they’re using it by virtue
of the set of rules that provide for their second speech, and to whom it’s
addressed.

That sort of thing can gives us who is addressed by the terms in a variety
of second utterances to the first speaker rules: Questions and answers.
Commands and returns. (Commands and returns very frequently have
misidentifications in them, e.g., “Get me a drink of water.” “Yes, Boss.”)
And then we’re in a position to consider what it is that the particular term is
doing - if we haven’t got, in the first instance, some good idea of what it is
doing.

In short, what you want to do is to treat the rules of conversation sequencing
as the first things to apply, before such things as ‘consistency’ and all the rest are
applied. And we can take it that before the second occurs, persons could say
who will talk, and who they will talk to. And that, again, gives us a way of
seeing how it is that the membership term is not doing that work.

Of course in this case there are a series of ways that the second utterance is
tied to the first. That is, both ‘yes’ and ‘thank you’ would lead us to suppose
it’s a tied utterance, even if we simply had the utterance itself and didn’t
know, for example, that an ‘insult’ had preceded, and the like. But of course
lacking the insult, it could perfectly well be, not a ‘second insult,’ but, for
example, an ‘appreciation:’ “Look what I brought you home.” “Thank you.”
The character of the whole item, then, what activity it’s doing, who is being
referred to, can be found, in part, via what the first was, since it can be a
second to a whole series of things. And of course the category name used,
‘Mommy,’ is to be found via both who it’s talking back to, and the kind of
activity it’s a return to, i.e., its character as a ‘second insult.’

I said that a ‘first insult’ can be followed by a ‘second.’ And now that thing
seems to be used regularly in a chained fashion, the second treated as a first
for a next, and so on, so that insults can be strung together. The game,
Tearing Down, is that sort of game, and there’s a literature on it. One place
you’ll find a discussion, but the name Tearing Down is not used there - it’s
an earlier study and it doesn’t deal with America - is The Lore and Language
of School Children by Peter and Iona Opie. There’s an article entitled
‘Ritualized verbal insults in white high school culture,’ Journal of American
Folklore , 78, page 337-344, and they cite an early literature on this.

What I want to do now is move through the various parts of this ‘tearing
down’ procedure. The next line is “Face the music.”

Roger : Ken, face it. You’re a poor little rich kid.

Ken : Yes Mommy. / / Thank you.

Roger : Face the music.

And that’s not a new insult, and as such, it seems that the fellow who
delivered the first is withdrawing, i.e., is not going to play a game - or to

Lecture 21

421

continue it, anyway. But someone else picks it up and offers him a teammate
in that game, against the other fellow. And he accepts the offer and begins
again.

Roger : Ken, face it. You're a poor little rich kid.

Ken : Yes Mommy. / / Thank you.

Roger: Face the music.

Al : Okay. Now you’ve got that er outta yer system. Now you’re a

poor little rich kid we’ve told you that.

Roger : And we also decided you’re a chicken shit.

The offer of a teammate is done in a rather nice way with “Now you’re a poor
little rich kid, we’ve told you that,’’ where “we’ve told you that’’ is affiliating
himself to that thing which the prior fellow said. And he’s thereby, I take it,
proposing himself as an ally. The prior fellow accepts that, in what could be
seen to be a collaborated utterance, “And we also decided you’re a chicken
shit.” Now, earlier in the session he had said, independently, “You’re a
chicken shit.” Saying it again here, in a collaborated utterance, what he’s
doing in part is saying, ‘You, having latched onto the last thing I said, and
having announced it for “us,” you might as well latch onto something I
earlier said. It’s now also for “us”.’ That is, since they’re making a team, they
might as well add up their resources.

We get a tremendously interesting line in return: “I decided that years ago
the hell with you.” The first part of it, “I decided that years ago” could look
like an ‘agreement’ and yet, at least for me, it doesn’t come off that way. And
furthermore, the second part, “the hell with you,” seems to be not consistent
with the first part as an ‘agreement’ (if you needed any reason to doubt that
“I decided that years ago” was an ‘agreement’). Of course the fact that it’s not
consistent with an ‘agreement’ isn’t necessarily an unequivocal matter, since
they may well have a series of inconsistent statements.

In any event, in dealing with “I decided that years ago the hell with you,”
which seems to be some kind of denial and not some kind of agreement, what
I wanted to see was how could it be a denial, and what kind of critique might
it be of the former remark? So the question is, can we establish that it is some
kind of critique, denial, or whatever? And how does that work?

The way I proceeded was to consider what kind of activity would it be to
decide that oneself was a chicken shit. We might say that it would be
courageous, i.e., it would take an act of courage to decide that you were a
coward. It’s facing a rather unpleasant fact, and decisions of that sort about
oneself are treatable as acts of courage.

Q: Isn’t it an assumption that such a decision is courageous? Such a
decision may not necessarily be an act of courage.

HS: I’ve been interested to explore what could be called the grammar or
the technique that persons use to arrive at and do the various things they do.
It has never been my aim to give medals to people for doing what they do.
And when I say that it’s an act of courage to arrive at the statement that

422

Part III

you’re a coward, I’m not saying 'This fellow’s done something courageous.’
Rather, is it the case that it could be said of somebody who says that they’re
a coward, that they’ve done something courageous? If so, that’s all the point
I need. And I take it that you think that that’s so, in that you’re saying that
it’s “not necessarily” the case that that’s so.

That’s all that I'm trying to propose. What can be said to be the
constructed procedures for arriving at an activity? What activity is involved in
arriving at another activity? In this case we have, deciding that you’re
something that’s a negative fact about yourself. And one thing about that is,
it can be something courageous. Certainly I have no idea whether this fellow
is a hero, or whether he in fact decided that years ago, or anything else of that
sort. For anybody who was to do such an activity, what makes it recognizable
as whatever it happens to be? That’s what we’re trying to deal with. And as
I have emphasized throughout, and I shall emphasize again, what I’m dealing
with are possibilities. And I’m trying to lay out the relationship between
possibilities. For some possibilities, if we get a possibility that we can
establish, that’s fine. If one is claiming that it could be another thing, first one
wants to show how it’s another possibility. That takes a task of analysis. Then
you may want to resolve the two possibilities, or the ten, or leave them open,
as you please.

Again, that he has done something courageous is something I have no idea
about, and I’m not making any statements about. If it’s the case that arriving
at such a conclusion as that you are a coward can be courageous, then we get
a rather interesting situation. And that is, that to say “I decided that years
ago” is to say something paradoxical. If it’s true that you decided you’re a
coward, then you’re not a coward.

Paradoxes of that sort - for which another term is ‘antinomies’ - could be
called ‘generative paradoxes’ in the sense that the procedure whereby
something is arrived at involves, for it to be arrived at, some other fact to be
present which, if it’s true, makes it false. Now the question is, do people in
fact play with paradoxes like that? Is there any evidence for supposing that
they do such a sneaky thing as that? And what gain is there in a case like this,
to make such a paradoxical statement?

The phenomenon of paradoxes has been a matter of very large interest in
such areas as philosophy and logic. The classical paradox is the famous
statement attributed to an ancient Cretan who said, “All Cretans are liars.”
If that is true, then it’s false, etc. And that paradox has had tremendous
influence in the history of philosophy and the history of logic. For one, it’s
been used within logic to formulate a basis for saying that no natural language
can be said to be consistent, and, relatedly, that the notion of truth cannot be
applied to a natural language, and a variety of other consequences which have
had as their import that logicians study the languages that logicians construct,
and propose that you can’t logically study natural language.

These inconsistencies which are developable within natural langauges have,
then, had very large import for logic. And as a consequence of their import
for logic, they have had an import for philosophy. There, the problem has

Lecture 21

423

been posed: Can you do philosophy in a natural language if it’s the case
that the natural language harbors inconsistencies? In such a language you
can’t, perhaps, then, decide that what you’re saying is consistent or true,
and therefore philosophy has to be done via the languages that the logicians
construct and study. In chapter 1 of any logic book, you can find
considerations of antinomies for logicians (for a fairly comprehensive review
of the literature on them, see the appendix to chapter 1 of Foundations of
Mathematical Logic by Haskell Curry. So the question of their import for
philosophy has been much argued about. Now, in an article called ‘Is
everyday language inconsistent?,’ Mind (1954)), A. Stroll is concerned to
argue that while it’s theoretically possible to build antinomies in English,
and probably in any other natural language, Members never do it, it’s a
resource that’s unexplored, and it’s therefore not to bother philosophers,
who can do philosophy, meaningfully, in natural languages.

Let me say that you can take it as a good bet that persons will do anything
a language permits. What we want to do is to see that, in fact, such
antinomies are used, and that they have a perfectly nice basis for being used.
First, however, let me give a quotation which depends for its sense on the fact
that it’s not heard as a paradox, and the work that’s done in hearing it
involves avoiding the possibility of paradox. This is from A Black Byzantium
by S.F. Nadel, page 71, note 3.

He’s talking about a tribe, the Nupe. The tribe was conquered some while
back by another bunch of people called the Falani. Now the Falani have been
absorbed by the Nupe, but they remain the ruling group in that place. So
anybody who is a Falani is also a Nupe, and of course some Nupes are
Falanis.

The following is a significant anecdote. The Etsu Nupe (Etsu is King)
once confided to me his grievance against a certain rather unruly relative
of his, whom he believed to have intrigued against him. He added, “All
Nupe are bad people, you know.’’

Where the sense of that depends on the fact that he’s treating himself as a
Falani for that remark, and not saying it about himself, but about his relative
and others. Nadel goes on:

The said relative, himself of course as much Falani as the Etsu, retaliated
by informing that “All Falani are liars.’’

For that remark, treating the other as a Falani, not himself.

The sense of that one depends on the avoidance of the possibility of its
paradoxical character, where that’s done via the manipulations of the
membership categories. Now I want to quote a clear use of the antinomy, and
perhaps we can get at what’s being done with the use of the antinomy in
general, and what’s being done in our case. This is from the New York Times ,
Sunday, December 5, 1965. It’s an article on Marc Chagall.

424

Part III

Between 1923 and 1939 many less perceptive critics and art lovers,
even in Paris, tended to confuse Chagall and Mane-Katz because of a
superficial similarity in their choice of Russian Jewish subjects. One day
in the early thirties, one of the Left Bank’s most notorious gossips asked
Mane-Katz what he thought of Chagall’s work. The answer came
without hesitation, “He is a great artist.” The same busybody then
hastened to ask Chagall what he thought of Mane-Katz. “He’s a very
bad painter,” Chagall replied. His opinion was promptly relayed to
Mane-Katz who shrugged his shoulders and (perhaps unconsciously
paraphrasing an ancient Cretan sophist’s example of paradoxical ambi-
guity) wittily disposed of the whole matter: “Well, all painters are
liars.”

What’s being done with the antimony there, and it’s a clear case, is that in
undercutting his own earlier statement he undercuts the other’s, as well. It’s
a very nice way, then, of presenting a remark which, if true, is false, and
thereby getting at the falsity of another remark, which would fall with it. It
does it without being a direct controversion of it; that is, he doesn’t argue by
saying, “No, I’m a very good painter,” but by producing a remark which is
false if true, so the falsity of the other is provided.

Then the issue is, not ‘why would a sentence be made, of itself,
paradoxical?’ but, what is it that can otherwise be done with a paradoxical
statement, i.e., it can perhaps undercut another troublesome remark. It’s
perfectly conceivable, I suppose, that the original “All Cretans are liars” had
a similar usage, though we don’t have its context.

So I take it that with “I decided that years ago” we have a quite elegant
return to the charge of cowardice. Of course it carries other possibilities with
it - for example, ‘It took me courage to decide that I was a coward. What
does it take you to decide that I’m a coward? Then, the end to it, “the hell
with you,” is perfectly coherent with the rest, being, possibly, an exemplifi-
cation of a piece of courage.

I want now to move to the second problem I posed, which was: How do
we provide for the transition between a ‘discussion’ and a ‘tearing down’? And
how do we provide for the order we find in their relationship, i.e., that
discussing comes first and tearing down second?

First of all, I take it that the fact that those two activities occur in a given
order - discussing first, tearing down second, is non-trivial, that it is a
recurrent order, and indeed, that given the first, the second is often expectable.
So it’s not simply that when the two occur they occur in that order, but that
if the first occurs, Members take it that the second can expectably occur. I take
it that that is clear. And let me point to certain problems which persons
regularly announce to each other. They can say, for example, “Why can’t we
ever have a peaceful discussion anymore? Whenever we get into a discussion
it ends up in name-calling.” Furthermore, when some topic gets started,
persons will regularly say, “Let’s not start on that, we’re liable to get into a
fight if we continue. ”

Lecture 21

425

Now what we can be doing in part is examining what lies at the basis lor
correctness, if we can see as correct those assertions that they make. But apart
from that, what we want to watch for is the kind of components that complex
activities have, and the ways that sets of complex activities can be put
together. Is that also orderly? If so, how? And is it also perhaps the case that
these complex activities are built by putting together in one or another
fashion, standardized pieces which can be used to put together other activities,
complex or simple? If so, an image I sometimes use, of people assembling
activities out of standardized parts, of a culture as a warehouse, can get a fairly
apparent sense to it. And that is the technical interest we have in the problem.

We have in the first instance, this phenomenon of the ‘discussion,’ which
can have a ‘topic’ - in this case, an ‘automobile discussion’ - and a variety of
positions taken on that topic, where as one good criterion for seeing the thing
as having been brought to a close, there is this business of arriving at ‘an
agreement. ’ There are a bunch of ways that an agreement can get arrived at
or not, but those are not what I’m concerned with now. I mention an
‘agreement’ because I want to differentiate it from the kind of criterion that’s
used in ‘tearing down.’ And also, to get at, in part, what provides for a
transition between discussion and tearing down.

Now we’ve got a set of positions offered. We could then have, as one
direction that the conversation goes, various attempts to get changes of
position, e.g., ‘You change yours, I’ll change mine,’ or ‘I’ll try to change yours
while retaining mine,’ etc. But there’s one possibility that is regularly used,
and that involves, for one, formulation by somebody of an explanation of
some other person’s position. And it may be a certain sort of explanation.
We’ve noted before that many activities are known to be category-bound, and
that positions can be category-bound phenomena. If that is the case, then one
sense of an explanation of a position is to see that the position is category-
bound, and to what category it’s bound. Now that kind of an explanation is
a fairly special one. It regularly constitutes an undercutting of the position that
was taken. That is to say, if you can find that category to which the position
is category-bound, you can use it as an explanation of the position, which
provides for the undercutting of the position.

And that possibility has had some rather major uses. One very prominent
sociology consists of the systematic exploration of that possibility for
undercutting all sociologies, including itself, and that is the sociology of Karl
Mannheim. What he was engaged in doing was to propose that all sociologies
(first he started with conservative sociology) were generated from some
position within the social structure, which explains how they came to say what
they say, and partially undercuts the objectivity of such positions. It’s another
case where some particular resource that Members routinely use, has been
turned into a more or less major scientific weapon. And it’s a weapon that
becomes especially powerful when it’s the case that those positions that are
used as accounts of what it is that was said, are constructed and controlled by
the sociologists that use them. That is, then, a much weaker check on
whatever kind of correctness there is for category-boundedness than there is

426

Part 111

when somebody says, “You say what you said because you’re a poor little rich
kid,” where anybody might well know that ‘poor little rich kids’ say such
things.

So in this case, asserting an explanation of some position via some
membership category, which accounts for that position having been taken is,
in part, part of the ‘discussion.’ It works in the discussion by perhaps
removing one of the positions from contention, and perhaps then, leaving the
other victor - if there’s only an other. However, it sets up its own
consequences, and that’s what I want to deal with, in part.

For one, there’s a considerable gain when an explanation via categorization
has been asserted, such as “Ken face it, you’re a poor little rich kid,” and that
is, that the criterion for success has changed very considerably. It can remain
‘agreement,’ but now on the question of whether he’s that sort of thing that’s
been proposed. And that’s not for him to decide, but a matter of agreement
among the others. What we get, then, is at least a partial narrowing of the
population in which agreement is to operate as a criterion of success. And in
this case, of course, we get the agreement by the third party.

For a great many positions taken in conversation, and especially - by
definition - those taken on ‘controversial’ topics - where ‘controversial’ topics
are defined in the first instance as those on which any given set of people can
be known to differ, and to differ by virtue of those categories they belong to
- it may well be the case that it’s an intrinsic possibility that at any point those
categories can be invoked, and, once invoked, the rest can be invoked.

Lecture 22

There is no lecture 22 in this edition; it has been combined with lecture 21.

427

Lecture 23

Agreement; What can be done with

?

0

I’ve been using, with regard to both the phenomena of ‘a discussion’ and
‘tearing down,’ the term ‘ agreement ,’ and I want to say a bit about it. It is used
within the conversation itself; one of the fellows says at one point, “Now well
I agree.’’ And there are other places where at least it might seem to be a way
that an assertion could be made, though it’s not. So, for example, after the
statement “Ken, face it. You’re a poor little rich kid,’’ we get “Now you’re
a poor little rich kid we’ve told you that,” where “we’ve told you that” is
kind of different, perhaps, than ‘I agree with what was previously said.’ As I
said earlier, “we’ve told you that” involves, it seems, formulating what was
previously said as having been said representatively, and therefore incidentally
said by the other guy, but said on behalf of us - even, perhaps, whether he
knew it or not. So that’s one sort of alternation to doing agreeing. And later
on in the conversation, at the end of the sequence of questions applied to the
entrant, there’s a bit of a discussion about whether he’s “just agreeing,” or
presumably whether what he’s saying is what he knows to be the case.

Now the activity of ‘agreeing’ is one to which some attention has been
given - much more than I can, and I only want to open it up as something
one might attend. One way that we get to open up such a phenomenon as an
activity, is to try to get some idea of it as something invented or discovered,
i.e. , a resource which may not always have been present. Here is a short citation
which kind of suggests that about ‘agreement.’ It’s by the classicist Gilbert
Murray, in a book of essays entitled Greek Studies (19 46), page 175. He says:

The early Greeks had, so to speak, discovered the logos. They had
discovered that often, instead of fighting you could say something, and
the thing said would make both sides agree. If people were bewildered
or puzzled they could say something and the thing said would make
them understand.

Presumably what he’s trying to get at there is, for one, that the resources of
a language - what it is that could be done with it - are things that get
discovered at some point, and when they are discovered, they may be
tremendous news. Where the activities that could be done with a language
may well be such kinds of things. In that regard, then, one might reconsider
an extremely famous formulation. It’s by Edward Sapir, in his book called
Language. This is from page 14.

428

Lecture 23

429

We’re not, in ordinary life, so much concerned with concepts as such, as
with concrete particularities and specific relations. When I say, for
instance, “I had a good breakfast this morning,” it is clear that I am not
in the throes of laborious thought; that what I have to transmit is hardly
more than a pleasurable memory symbolically rendered in the grooves
of habitual expression. Each element in the sentence defines a separate
concept or conceptual relation or both combined, but the sentence as a
whole has no conceptual significance whatever.

It is somewhat as though a generator capable of generating enough
power to run an elevator were operated almost exclusively to feed an
electric doorbell. The parallel is more suggestive than it at first sight
appears. Language may be looked upon as an instrument capable of
running a gamut of psychic uses.

Consider, in the first instance what, for him, is obvious; and that is, that what
it is that could be said to be being done by use of language specifically is
transparently obvious, stateable by stating the sentence itself (‘‘I had a good
breakfast this morning”), and then to be seen to be using very little of what
language really could do. What the import might be for researchers, of a
statement like Sapir’s - small or large - might turn on whether, for example,
our understanding of the statement is invoked in the same way he constructed
it, or whether it was taken to pose some more or less significant problems.

I’m going to quote another passage that refers to this earlier one, and which
does draw some consequences from it. This is from a paper, ‘The semantic
structure of language’ by Uriel Weinreich, in Joseph Greenberg (ed.),
Universals of Language, pages 117-118.

In a remarkable passage, Sapir likens language to a dynamo capable of
powering an elevator but ordinarily operating to feed an electric
doorbell. Language is used more often than not in ways that do not
draw upon its full semantic capacity. In its ‘phatic’ functions, when
speech is used merely to signify the presence of a sympathetic inter-
locutor, it easily becomes ‘desemanticized’ to a formidable extent. In its
various ceremonial functions, language may come to be desemanticized
by still another mechanism . . .

The more pressing task for linguistics, it seems to me, is to explain
the elevator, not the doorbell; avoiding samples of excessively casual or
ceremonial speech, to examine language under conditions of its full-
fledged utilization.

Now the last thing that I intend to be doing is legislating for linguistics.
But the idea is held that we know, right off, where language is deep and
interesting; that we can know that without an analysis of what it is that it
might be doing. One wants to see, for one, that the set of ways a statement
like “I had a good breakfast this morning” can be talked about; for example,
‘phatic,’ ‘conceptually rich,’ ‘conceptually poor,’ ‘non-conceptual,’ etc. are all

430

Part III

given as a set of alternatives, again without it, or anything else being studied
very much. But a program is laid out. And among the important things about
that program are what’s not to be studied. And what’s not to be studied is
proposed, not by virtue of, for example, that today we couldn’t study that,
but by virtue of presumed results. Those are really quite extraordinary
arguments. And if they don’t proceed quite strictly from a notion that we can
or must know what the aims of a discipline are before we begin, then a good
deal of that is involved in such formulations.

Such questions as what language can do, what people can do with
language, what the results of an analysis of “I had a good breakfast this
morning” would involve, what kind of program it poses for a field, all these
things remain absolutely open.

Indeed, more remains open, since we might ask, for example, by what right
(assuming Sapir to have heard that sentence) could he have proceeded to pose
it as a singular problem, i.e., to have isolated it from whatever it happened
to occur within, and to suppose that it could be analyzed, transparently, in
that fashion. For one, at the root of several fields concerned with language and
with much other social behavior, is the idea that you can pick out an item,
however you want to, and consider that the isolated item contains the
properties that it would, must, ought to have, when it otherwise occurs. But
one would, for example, imagine that the sentence “I had a good breakfast this
morning” was not the first and only sentence uttered in the conversation.
Maybe it was the answer to a question. Maybe it came well into a conversation.
Is it the case that anything occurring anywhere in any conversation can be
isolated without regard to the fact that it occurred in such a place? Is it the case
that a language is built to provide that that should take place, and to allow for
stability of things like meaning, etc.? What’s being done with this thing?

I’ve gone through this discussion, for one, to notice that it is perhaps not
incidental that people have not devoted their life to studying sentences like
“The baby cried” or ‘‘Hi, how are you.” There may be more or less defensible
reasons for not studying such sentences. Not doing that, however, has real
consequences.

I started out by noticing that what it is that can be done with language, i.e.,
the resources of a language, are things that get discovered at some point. Such
ideas as that we already know where language is deep and interesting, and
what it is that it might be doing, might lead us not to see that some things
are properly - at least for heuristic purposes - to be seen as inventions at some
time. Now in the case of ‘agreement,’ certain aspects of its use can be said to
have been great discoveries. So, for example, one of the things that’s said in
the history of religion is that one of the great contributions of the Jews to the
history of Western religions is the fact that they were the first to formulate
their relationship to God as a matter of ‘agreement,’ where that involved an
enormous shift in conceptions of religion.

That matter posed a long bit of historical research into the question of
where did they come up with the idea of a covenant binding them and God?
To which, about ten years ago, a proposed solution was offered. A fellow

Lecture 23

431

named Mendenhall, a student of the history of religion at Michigan, was
working in the Hittite archives. The Hittites were a people who controlled the
ancient Near East some 3,000 years ago. And they had massive archives
which were discovered toward the end of the 19th century, which people have
been working in ever since. And there, Mendenhall found among some of the
archives, what turned out to be the Hittite vasselage agreements. That is to
say, those agreements which the Hittites made with the various small tribes
that they controlled. And what he found was that essentially nine of the Ten
Commandments could be found as parts of the normal Hittite vasselage
agreements, which ran to the effect that ‘We are your only ruler,’ etc., just as
do the Ten Commandments. That is, of course, only proposing that that’s
where it came from; how it was transformed is another issue. And more
recently it’s been criticized somewhat. But the possibility of the political
derivation of such a phenomenon is of some interest.

Now, the recognition that one can do the activity ‘agreeing’ is treated as
something that the ancient Greeks discovered, though lots of other peoples
would have discovered it as well, presumably independently. And of course
the question of independent discovery of this or that resource is a nice one. It’s
said, for example, that science as we know it is something that only the Greeks
discovered independently, and everybody else learned it from them. And
what activities are discovered by every culture independently, then, gives us
some idea of more or less intrinsic properties of, say, a language.

Agreement can be of great interest by virtue of the things that seem to be
involved in it. For one, the notion of the comparability of statements - and
that, of course, is fundamental to the development of logic, mathematics, and
the rest. Now it’s not simply the notion of the comparability of sentences,
because presumably it was discovered that even sentences which were not
exactly alike were ‘the same.’ That’s the sense of making a statement in a
sentence. That this comparability could be used for purposes of deciding
‘sameness’ across persons, across time, without respect, perhaps to some
extent, of context, that would be quite an achievement. And it would be a
legislated achievement, where presumably at least for theoretical purposes, it
is just as viable to say that two statements, even if they’re ‘exactly the same
sentences,’ occuring at different times or produced by different people, are
different. And then perhaps we can get at what are some of the viable notions
of ‘sameness’ for activities, via a consideration of what it is that can be said to
be ‘one statement agreeing with another.’

Now, for Members, ‘making a statement' is seen to be something different
than simply uttering a sentence; that is, it becomes an activity for which
language is used, but which could be done in other ways than the particular
way it was in fact done, and still be seen as ‘the same thing.’ You could say
it differently and it would still be the same statement. And furthermore, the
activity of building it could have been directed to ‘producing a statement that
is the same as some other statement that some other made.’ And it could be
recognized as that, which would then involve seeing that a statement had been
made that was ‘the same’ as yours, without regard to the fact that it was made

432

Part III

by somebody else. And that, then, involves some notion of studying actions,
and leaving the actor in it as somebody who can be seen to be irrelevant, for
some purposes at least.

Again, a notion that statements can be ‘the same’ though the sentences are
different is by no means an obvious one. And of course the essence of the
notion of ‘ceremony’ often is that exactly the statement - whatever that might
mean - has to be redone or failure will occur, if the statement is powerful.
Where, then, memory is the crucial matter, and memory has to be helped by
very special techniques - most particularly, the use of poetry as a mnemonic
device.

The forms of poetry become the ways that information is captured so as to
be reassertable in just the fashion it was first asserted. And that is the basis of
Plato’s attack on Homer. In the Republic , where Plato attempts to reformu-
late the problem of knowledge for the Greeks, he launches a very concerted
attack on Homer, proposing that Homer is the storehouse of knowledge for
the Greeks, but it’s lousy knowledge. Its effectiveness, however, is given by
the fact that it just gets built into everybody’s memory via the poetry, and if
you’re going to have a progressive knowledge, you have to undercut that
form. That issue involved in Plato’s attack on Homer has been recently dealt
with, in some detail, in a book about oral knowledge called A Preface to Plato
by Eric Havelock (Harvard University Press), which is much worth looking
at.

In any event, for those kinds of things it’s taken to be that the statements
are powerful in some way or another, and if they’re followed exactly, then
some set of consequences will occur. And for matters that were so frozen,
‘agreement’ has to be irrelevant, and the notion of ‘agreement’ might never
even emerge - it’s simply an issue of repetition.

Now the emergence of a notion of ‘agreement’ as a possible thing that can
be sought and dealt with, is, for purposes of available history, to be marked
as an invention of Socrates, I suppose, for what was characteristic of his
technique was that he was willing to start with any statement that the person
he was talking to held, and then bring him to see that that statement and
some other statement that he would make were inconsistent, and therefore
that there was something wrong. It seems absolutely obvious to us that if they
are inconsistent there is something wrong, but presumably that had to be
learned as a complaint. If you took it that what was said yesterday need have
no relationship to what was said today, what was said in one context need
have no relationship to what was said in another, and in any event that
persons were well formed and ethics were consistent - whatever they were,
they were consistent - then no such attack as that which Socrates launched,
could be made.

Now Nietzsche in his fantastic attack on Socrates - a chapter in, I think, his
book on Wagner, called ‘An attack on Socrates’ - sees just what it is that
Socrates is up to, and treats him as destroying the Greek ethic; an ethic which
didn’t involve people having to know what their views were, they just ‘acted’
as he puts it, and they didn’t have to have answers to questions as to what

Lecture 23 433

their opinions were. And when they had to, the society became sick, and
subject to the kind of undermining that Socrates wrought.

This phenomenon of ‘inconsistency’ that Socrates used, is a criterion which
can have a kind of a strong appropriateness for a discussion, but it was
transformed into a strong criterion of a total ethic. That is to say, one was to
live one’s life with a notion of ‘consistency’ operating for any set of activities,
where, then, the sets of activities become things that are routinely transferable
into statements, and the statements are to be compared, and by comparison
of the statements one can decide their consistency. And if that’s the use, then
one can begin to see what relationship there is between, as he claims,
knowledge and virtue. One was bound to be un virtuous if one did not
produce a set of activities which, once transformed into a set of statements,
would be found to be consistent. And it’s too much to expect that without a
statement, i.e., a set of rules, which one used in generating any given activity,
one would build consistent activities.

The notion of ‘inconsistency’ has become enormously powerful, and is now
used as an explanation for all sorts of troubles, leaving completely aside the
issue of its use in, say, logic. One gets theories of psychopathology which turn
specifically on the presence of inconsistency; where, for one, the project of
some psychotherapy is the production of consistent activities. And these can
get very fine. The Bateson ‘double bind’ theory of schizophrenia proposes, for
one, that the basis of schizophrenia - or that schizophrenia that they study,
anyway - is that the pre-patient is subjected to at-the-same-time-inconsistent
activites. So that, for example, in a statement that the mother makes that’s
invitational, she also engages in activities that are rejectional. In an activity she
does that’s loving, she also does an activity that’s hating. This, then, is
supposed to generate confusion and the rest, and produce schizophrenia.
Where what has to be core to that is the effectiveness of consistency or
inconsistency - and that, of course, is a historical invention.

It’s a very magical idea, agreement. And it’s treated as magical ideas are
treated, i.e., wishfully: ‘If we could only agree then everything would be
alright. ’ And how it is that, agreeings having been reached, the world is then
relaunched on a solid basis is not really enormously clear, but involved in it
are those notions that also give rise to the hopes for language reform and what
it would do: ‘If only we had better communications,’ ‘If only we used, not a
natural language which is heir to all the historical difficulties that it has, but
some language constructed by logicians or proto-logicians, then the great
range of troubles that we find in our world would disappear.’ There’s
tremendous hopes involved in what it is that language can do, which are
partially involved in overcoming what it is that language does.

I suppose pretty much any conflict that we have eventually gets formulated
as one that discussion could remove, where it could remove it if discussion
could produce agreement. The set of projects that have been generated under
such a hope include things like the United Nations, and so on. And we could
ask what would be some basis for any such hope being entertained. It could
hardly be Socrates’ successes, because after all he was forced to take poison in

434

Part III

the end. And if it’s such things as the fact that occasionally in a household, if
there’s a dispute as to when it is that one should go to sleep, or what kind of
car one should buy, or the like, discussion can resolve it, you’re getting a
rather fantastic extension. But then, discussion is something that everybody
knows about, and agreeing is something that everybody knows about, and
perhaps it is, then, a kind of institution which can appeal by virtue of its
familiarity, in the sense that Christ appealed with love as something that
everybody ought to know about, and hoped also to change the world.

Now it might appear that the focus I’ve given to conversation would
involve me in saying that the rules of conversation are to be given such an
enormous status. I point these things out to indicate that they seem to be
given that status without reference to whether they can bear it. There are,
then, a whole set of properties of language that have had some success, more
or less limited, more or less great, for language, that are used to handle an
enormous range of other things, presumably, in part, by virtue of the fact that
those other things can be talked of, anyway, and therefore can permit an
extension of the talk of them to those things that are talked of.

Methodologically, ‘agreeing’ is a tremendously troublesome thing to deal
with, in that one needs, I guess, not merely a notion of ‘same’ for those things
about which the agreements can be, but a notion of ‘same’ for the activities
of agreeing. And that matter is normally simply supposed away. The formula
for its being supposed away are things like social contract theories, conven-
tionality bases of everything - or some things, such practical tests as, ‘If it
turns out to work then they must agree,’ ‘they can talk so they must agree,’
at least on agreement, anyway. It is supposed away in almost any public
opinion survey which would say ‘Do you agree on X,’ assuming that
whatever it is that ‘X’ is may be problematic, but whatever it is that
‘agreement’ is would always be treatable as ‘the same’ for such purposes as
counting, anyway. And if it varies, then you could write ‘strongly agree’ or
‘weakly agree,’ or whatever else.

Now what other properties are involved in the use of the notion of
‘agreement’ than those few I’ve mentioned, I could hardly say, and certainly
we have to know. And what things are demonstrably doable for some aim
like ‘getting knowledge,’ ‘getting a piece of work done,’ ‘getting a proposal
through,’ or whatever else, with all those aims replaced by ‘getting some
agreement’ on them, is another thing that has got to be examined, for the use
of that image would be a basis for having technical specialists in ‘agreeing’
brought into any given problem to help in its resolution. And that, of course,
is a good deal of the basis for the omni-presence of lawyers. That’s what
they’re specialists in. The effectiveness of the use of people like that is another
question.

There is, in short, no institution, I take it, that has had the hopes laid on
it that ‘agreement’ has, since whenever it was discovered as something that
could occur in conversation. Whether it can bear those hopes, and how it
could bear those hopes are matters that, I suppose, require some rather direct
attention.

Lecture 24
Measurement systems

Roger : Well first of all you must be crazy or you wouldn’t be here.

Ken : heh heh

Jim : Yeah that’s how it goes.

Al : Secondly, you must be an under-achiever.

Jim : Yeah,

Roger : You hate your mom and dad huh

At : -third of all.

Jim : Oh, somethimes.

Al : Fourth, you like to drive cars fast.

I’m going to talk about the phenomenon that we can consider via this
statement, “you like to drive cars fast,” most particularly focussing on “drive
cars fast. ’ ’ There is a related discussion on the first page of the data handout . 1

Now, there is a large issue lying behind an investigation of something like
this phenomenon, i.e., talk of how one drives cars. If it is the case that there
are alternative measuring systems available at any time when some new
phenomenon is introduced into a culture, which the culture is measuring, then
what is it that serves to control the selection of some measuring device from
among the alternatives? And then, relatedly, what are the features we ought
to use to compare measurement devices? The case of cars can be a fairly nice

1 Following is a simplified version of that segment.

Roger : I have got a speeding ticket on the average of one every three and a half

months. Since I’ve been driving. And I deserve it.

Ther : ha ha ha

Roger : I deserve it. Because I always drive about seven miles an hour faster

than everybody else. You know? Like everybody’s driving thirty five
I’m driving about forty two somethin’ like that. So this is the
conseque(hh)ence I have to pay. ‘Cause I can’t stand driving that slow.
Specially when I think I can handle a car at that speed. In traffic.

Ther : Uh huh

Ken : Oh I can think of a million times when I should’ve gotten a ticket, going

on the freeway. I’m always goin’ seventy five. I mean I’m just I’m not
racing, I’m right along with- or I’m right behind the guy ahead of me
and he’s just he’s goin’ as fast as I am and lotta people pass us up like
we’re sittin’ still.

435

436

Van III

one because they are relatively new, and when they were introduced there
were, obviously, a variety of measurement systems available. Furthermore,
built into the car is one possible take-off device for measuring, i.e., the
speedometer.

We could pose those problems directly, by virtue of consideration of how
it is that measurement devices are used, and what seem to be their different
features, rather than, for example, simply supposing that we ought to use the
available mathematical considerations for deciding such issues, e.g., that the
terms are ‘ordinal’ versus ‘cardinal’ or whatever else.

There is, furthermore, an issue which concerns the theory of ethnography
involved in consideration of measuring systems. That arises in the following
way: What is it that an ethnographer uses as some sort of base - if he’s going
to use a base - for constructing a formulation of some measuring system used
by a tribe? The overwhelming tendency is to use some ‘real’ - that is, defined
as scientifically correct - system, in terms of which one can formulate how it
is that the natives go about measuring whatever it is that they’re measuring.
So, for example, color terms may be mapped onto the ‘real’ spectrum known
by Western science. And one can, in that way, construct formulations of
native color systems. It tends to be the case that even those anthropologists
who consider themselves most radically oriented to an attempt to describe
native systems of categorization - native thought systems - find themselves
using their formulation of the real world as a way of getting a hold on what
the natives are doing.

Now when one starts with a line like “drive cars fast,” one might suppose
that one is right off dealing with a term from perhaps a set of terms that are
in one or another obvious ways, more or less vague. ‘Vague’ by contrast to,
even for the natives, an available clear way of formulating what they’re doing.
That is to say, one might, in parallel to a consideration of, let’s say, color
terms, put a speedometer on the blackboard and map onto it the system of
terms used, and say, then, that for Americans, or California kids, or whatever,
‘fast’ means from point one to point two, ‘slow’ means from point two to
point three, etc., where you would say that ‘vagueness’ arises at certain places.
It’s not clear whether 50 is ‘fast’ but it is clear that 90 is ‘fast.’ It’s not clear
whether 30 is ‘slow,’ but it is clear that 15 is ‘slow.’

And you could construct a testing procedure, which would have partici-
pants telling you which are ‘fast’ and which are ‘slow’ positions on the
speedometer. And that might be a way for you to decide, for example, not
simply that they use a stable set of terms, but whether those terms are
psychologically real or not. So, for example, it is proposed as a development
in, say, componential analysis, that one can or cannot construct psychological
dimensions to the various terms that they analyze, and that one can see
whether those terms have a psychological reality via some such test as I’ve
skteched.

There are serious problems with any such conception. The ways that one
might procede to get at them would involve a questioning of the notion that
one’s procedures are in the first instance to hit some base which the natives

Lecture 24

437

must, after all, be coming to terms with. Where essentially their activities are
to be seen as ‘coming to terms with some phenomenon,’ such that one could
then explain, for example, the defects of their system, by virtue of other
interests that they may have, or by virtue of the simple inadequacy of their
conceptual systems - a matter which is obvious enough, they’re primitive.

If, however, we seek first to take a set of terms and see how they work,
examine some of their virtues, perhaps, then we might be able to work out
what are the criteria for usable measurement systems. And then set up a way
of comparing alternatives to decide which ones are ‘better,’ and arrive at some
way of seeing how come they selected one or another. One might not, then,
arrive at the conclusion which sees a system in which ‘fast,’ say, is a term, as
being more or less defective.

And it’s in such a kind of general light that one would come to look at
what is otherwise, I suppose, a rather dull remark, “You like to drive cars
fast,’’ which might also seem analytically opaque, i.e., what could you do
with it?

The set of terms of which it is a member seem to be extremely simple, and
the task doesn’t much involve the location of them, but consideration of them
once you have them. Essentially there seem to be three: ‘Fast,’ ‘slow,’ and
‘with the traffic.’ They have some extraordinary virtuosity of use, which
virtuosity seems to be involved in the fact that they are used, and those are
things which we might just begin to pick off - where, among other things we
want to get to eventually, are considerations of what is it that provides for
their usability, as an issue. Now that problem is kind of a neat one, and it’s
one that you might not get at if you started with a speedometer.

For example, if you started with a color spectrum as what had to be come
to terms with, you might not easily come to a direct consideration of the
special usability of the set of terms that were in fact constructed, and,
furthermore, the special considerations that permit their use. For the
considerations in the case of the color spectrum would essentially be that the
physical world was so constructed, without regard to some given tribe’s
problems. And, at least in the case of these speed terms, that’s not the issue
at all.

The core thing about their usability, for one, is that persons can be seen to
clump their cars into something that is ‘a traffic,’ pretty much wherever,
whenever, whoever it is that’s driving. That exists as a social fact, a thing
which drivers do, where it’s when that isn’t usable that you have some trouble
seeing how you’re driving, i.e., on roads where there are no other cars, or only
you and another, perhaps. And by ‘a traffic’ I don’t mean that there are some
cars, but there is a set of cars that can be used as ‘the traffic,’ however it’s
going; those cars that are clumped. And it is in terms of ‘the traffic’ that you
see you’re driving fast or slow.

But it’s not the business of anybody to set themselves up as a kind of a
culture hero because they make it their business to be ‘the traffic’ wherever
they happen to be, so as to allow others to orient to them. I say that not
without some irony, because Robert Merton would have them be a kind of

438

Part III

culture hero; they are the people who make it such that everybody else can do
deviant activities. That is, you can imagine those people who are driving in a
clump as permitting the measurement system to work, taking on the
character that they sacrifice themselves to others’ pleasure, etc. Like the
middle class would talk about their existence as providing those things which
Beatnicks can deviate from; having, then, a kind of heroic status. They
sacrifice themselves to permit a set of facts to occur, which permit others then
to see how they can have a good time violating them. It’s a perfectly
legitimate way of thinking of themselves.

Again, the core point being that it’s not a law of nature that there is ‘a
traffic. ’ Some people go about driving in such a way that permits themselves
and others to see that the traffic is going at whatever rate it’s going, in terms
of which some can go ten miles an hour faster, or whatever. As compared to
the fact that it is a law of nature that the spectrum has certain properties.

Let’s consider some of the features of this system. First of all, it’s not only
the case that you can at any given point decide that you’re driving fast or slow
or with the traffic, but the same sets of terms are used to formulate how you
drive in general. Secondly, that set of terms permits you to decide how
anybody in your view is driving, on any given occasion: He’s driving fast, he’s
driving slow, he drives like us.

That’s one set of things. But what we also want to see is that the
comparisons that are set up can be used across whatever driving conditions
there are, and whatever speed laws there are, such that two drivers can talk
to each other, consider how it is that each drives, where they don’t know, for
example, where the other is from, what the speed laws are, what the normal
conditions of traffic are, etc. “How do you drive?” “I drive fast.” “I drive fast,
also.” Where what the conditions are under which you drive, and what speed
you drive at when you’re ‘driving fast,’ are irrelevant.

Now that means, for one, that one tends to see that the fact of ‘a traffic’ is
not simply an incidental occasion for the piece of road on which you now are,
or the piece of road over which you’ve traveled and are likely to travel later,
but that wherever people drive, they drive that way, such that, for example,
you can move from place to place over time and find yourself in a setting
which you can orient to as having those features. The stability of the terms
and the conditions under which they’re usable, are then such that time, place,
speed laws, whatever else, are all irrelevant to their use. Changes in speed
laws, changes in the capacity of the car, changes in personnel, new
generations, new places - this thing can hold.

We might consider what would be involved were one to have replaced
these terms with a set of numbers (“Now I’m driving at 37 miles an hour”
or “. . .40 miles an hour”) and consider the conditions under which a set of
numbers might be usable to do the range of tasks that this system is usable
for. For one, perhaps, one would have some rather complicated problems in
arriving at a notion of how one drives. Would you do a random sampling
over a year? Would you have to throw out your past experience as cars
change? As the laws change? Would you try to consider the relevance of the

Lecture 24

439

fact that you regularly drive for some part of your time through school
districts? On freeways? How could you talk to another? What kind of
information would be required to find out what life is like for him? And there
is no intrinsic reason why one shouldn’t see a set of positioned terms as ‘clean.’
Except by virtue of the fact that one takes it that a set of relative numbers
defines what would be ‘clean,’ and that the terms are positioned via the
numbers, rather than that they are positioned with respect to each other.

For all we know, with this system, a term once having been arrived at,
whatever it is that happens in the future, it will be usable. It might, however,
provide for some real difficulties, e.g., for those people who ‘always drive
fast,’ that might get them into trouble insofar as they do regularly orient to
the traffic to decide how fast they’re going to drive, i.e., faster than it. Then,
of course, the traffic can control the kinds of deviations that occur.

In any event, there is a kind of extraordinary adaptability to the use of this
system; such matters as having a set of terms which are usable at any time,
where the same set of terms is usable to formulate how one characteristically
proceeds, and furthermore, can be used to compare how one is driving now
with how one characteristically proceeds. To compare, also, oneself with
anybody else. And to do that comparison without any concern for the
particular speeds, conditions, etc., under which the system is used.

What we want to ask, then, is whether such virtues - if they be virtues -
are important to the selection of some system, and the extent to which what’s
important is that there be some set of persons who make ‘the traffic,’ and who
make it as an unassigned task. In that regard, it’s curious that you can
regularly tell, as you move up to a clump of cars, for example, that somebody
is making the traffic as a task. You can tell from the rate at which ‘the traffic
ahead of you is moving’ that, for example, it’s being made by a cop sitting
in front of it.

Another question is, can a thing like this be used if it’s only partially used?
That is to say, if not pretty much anybody who learns to drive learns to use
it? Where, if for some measurement system it’s important that pretty much
everybody who does the activity for which it’s relevant uses it, that’s another
set of constraints; a set of constraints which such a system may more easily
satisfy than, say, speedometer readings and the kinds of comparisons they set
up - unless, of course, it were the case that, for example, there were no
variation in speed laws, and that the speed laws were taken to define the point
one should arrive at, i.e., if the sign said ‘40 miles an hour,’ then one speeded
up until one got to 40 and held it at that.

Now if a system like this is used, and those people who deal with the
system - who enforce the laws and deal with accidents and the like - don’t
really attend to it very carefully, it may have a range of consequences, itself.
So that, for example, it may be no solution to the problem of fast drivers and
the troubles they get into with others, to raise or lower the speed limits.
Furthermore, in a situation where there was any kind of occasional variation
in the rate of speed, like an accident, like a road being washed out, like a
highway being repaired, there might be tremendous troubles if the given

440

Part III

speed that was posted were treated as the only norm one could orient to. But
‘the traffic’ can adapt to all such things, and ‘fast’ or ‘slow’ drivers can adapt
to ‘the traffic.’

It is, I suppose, a rather lovely problem for those persons who like to
address largely the problem of social order, to ask what would ( )

do with the phenomenon of ‘the traffic.’ Or what would other theorists do
with it? How would they explain it? Would they find it tremendously
puzzling, i.e., everybody ought to be driving fast, if fast is something that
permits expression of, say, strength, or whatever else?

Again, as I mentioned, it would be via such a system’s use that we do get
the troubles that we apparently do get; for example, people driving across the
country, when they go through Montana where there ‘is no traffic,’ find
themselves in a kind of situation of sensory deprivation, not knowing exactly
what to do, how to deal with the occasional car, or the kind of misery one feels
when a car one has been using decides to leave the road. That is, the sorts of
things that would be more or less irrelevant, were it the case that one had as
one’s trusty guide a speedometer, and that’s it.

Lecture 25 [Incomplete]
‘Company’ as an alternation category

At : Now let’s see what else can we decide about / / you.

Roger : Hey don’t tear him down.

At : : We’re - Ok

Ken : I’ve been / / torn down // for ( )—

A/: ( // )

Roger : We got company heh

At : (hh)Oh Ok.

I want to talk a bit now about this line “We got company.” It’s the first
assignment of a category to this fellow since he’s come, and what we want to
consider is the sort of category ‘company’ is, and what’s involved, perhaps, in
its use here and now (that is to say, there and then).

The category is an alternation category to a whole series of membership
categories, each of which has it as an alternation category under some
circumstances, i.e., there could be ‘family ’/‘company’, etc. But wherever it’s
used, it is used as a category in alternation to some group. And there are
certain things in common in its use, and for those terms that it stands in
alternation to. For example, I think it’s the case that what it stands in
alternation to is regularly something which is territorially constituted. So that
if the three boys and this fellow were co-present elsewhere, they might not
talk of him as ‘company.’ And, for example, it is by virtue of somebody
arriving at a family’s house that they may be seen to be ‘company,’ whereas
in other circumstances his co-presence is not that sort of thing.

Of course for some of the categories that have ‘company’ as an alternation,
the category is usable when the alternation category is not, i.e., a family in
some places other than their house is still ‘a family.’ But this set of persons
may not be ‘the group’ in other places and times. To use that term ‘company,’
then, can be a way of invoking the fact that at this place and time, they are
constituted as something that has proper members.

Now for “We got company” there’s a question of who ‘we’ is. And it can,
perhaps, get a fairly decent solution by virtue of the fact that ‘company’ is
used in alternation to it, where we can ask who is it that might not be
permitted to say they “got company.” I suppose, obviously, the therapist.
What he “got” is a patient. ‘We’ then, may refer to the members of the
group; i.e., the patients other than this new one. And if it’s the case that the
therapist might not properly refer to this fellow as ‘company,’ then the fact
that they do so refer to him has some real interest. We can come to ask what

441

442

Part III

uses ‘company’ has for them, and consider the other things that they could
refer to this fellow with, and what they might involve.

‘Company’ is an alternation category. What, then, might be, for example,
other alternation categories, from which one was selected - this one. And
then, furthermore, what is to be said about the fact that any alternation
category was chosen, since it’s at least plausible that under some circumstances
they might not be in a position to use an alternation category, but, given the
fact that the therapist has selected a new patient, he might, by that alone, be
‘a member of the therapy group,’ not only from the point of view of the
therapist, but from their own. That being possible, the sheer fact of an
alternation category being used gets a clear relevance. { tape cuts off }

Lecture 26

Being ‘chicken’’ versus ‘giving lip

back’

Al : Tell us all about yourself so we can find something bad about

you.

Ken : Yes, hurry up.

Roger: Well first of all you must be crazy or you wouldn’t be here.

Ken : heh heh

Jim : Yeah that’s how it goes.

Al : Secondly, you must be an under-achiever.

Jim : Yeah,

Roger : You hate your mom and dad huh

Al : - third of all.

Jim : Oh, sometimes.

Al : Fourth, you like to drive cars fast.

Jim : Yeah,

Al : Fifth you like to ta - you like uh wild times,

Jim : Mm hm,

Ken : He smokes like / / me, see?

Al : Sixth you like booze,

Jim : Huh? / / Yeah

Al : Seven you like to smoke

Jim : Mm hm

Roger : And seven you’ve been arrested for rape and other things.

Al : Eight you heh

Jim : No, not that,

r Ken : hehheh

-Jim : hehh You’re gonna ta(hh)lk

L Al : Eight you - Eight you uhm give lip back to everybody

Jim : Yeah,

I’m going to talk this time about the line, . . you give lip back to
everybody.” Let me give it some context. One of the things I think it’s
relevant to, and what, in part, I’ll be directing my remarks to, is the statement
that starts off the sequence of questions (or comments, or remarks, as they
may be) in the group therapy session, ‘‘Tell us all about yourself so we can
find something bad about you. ’ ’ I want eventually to be able to see if we can
come up with an analysis which would show how such a statement could be
made to anybody. That is to say, that one could be able to say to anybody that

443

444

Part III

if you ask enough, and they say enough, then something bad about them will
be found.

In constructing a solution to that, I want to be able to do it in what would
be the most interesting way. And what I’ll be doing, then, is building an
analysis which is not anywhere near as general as we need, but which may deal
with matters that would be involved in a more general solution.

What would seem to be involved in a general solution is a situation in
which there is a dilemma which can be formally arrived at, and where - if
there are two choices - either choice can, under one or another formulation,
permit the statement ‘you’ve done something bad.’ If we can find that there
are some general dilemmas which can be formulated as holding for
anybody - under some formulation of ‘anybody’-dilemmas having this
characteristic that either alternative can be seen to be ‘something bad,’ then we
would have a nice way of showing that the statement “Tell us all about
yourself so we can find something bad about you’’ could be shown to be true.
The dilemma itself, of course, would have to involve some action-alternative
which everyone would face. So we need a dilemma like that, which we don’t
have right now.

So one of the first tasks is to get us the sort of dilemma that this statement
names a solution to. And that’s what I’ll try to talk about now. The first thing
we want to notice is the following. We could have a reading on “. . . you give
lip back to everybody’’ which involves, let’s say, taking each term one by one,
finding what it means, and then seeing whether the statement adds up. So
that, for example, taking ‘everybody’ independently of what’s gone before,
we could say the statement’s absurd, simply by virtue of the fact that we can
suppose that he hasn’t talked to ‘everybody.’ The question is, is that the way
to find out what ‘everybody’ refers to?

What I want to propose is that such an idea would involve a possible
proper sense for ‘everybody’ if, for example, we had a statement like “You say
‘hello’ to everybody,” where what’s involved in that is that the action you do
is the same action - saying “hello” - to whomever you do it. Whereas I think
it’s the case that whatever it is you might do - say, being insolent - it would
only be ‘giving lip back’ to a special class of people. Then, given the fact that
you could only do ‘giving lip back’ to a special class of people, it might be that
you do give it to all of those people, on the occasions under which you can. So
‘everybody,’ then, would have a rather more restricted sense than it could
have independently, by virtue of the fact that there are only some people to
whom what you’re doing could be considered ‘giving lip.’ And there’s a brief
characterization of a possibility, earlier in the session, on the second page of
the data handout. They’re discussing cops, and Roger says:

Roger: He can bust ya on a thousand things non-moving violations if

you give him any lip, even if you didn’t do anything wrong. They
can just go over your car and write you up ticket this long.

It’s in the first instance kind of curious that on a variety of references to this

Lecture 26

445

phenomenon ‘giving lip back,’ extremely similar sorts of events are pointed to
by kids as occasions under which it could occur. And apparently it is
something that they all take as a possibility, so that if, to ‘ ‘you give lip back
to everybody,” he’d said “no” instead of “yes,” then, I take it, it wouldn’t be
heard as saying, for example, that he’s never had the occasion to do it, but
that having had the occasions, he didn’t.

Now the question is, can we formulate some class, of which this fellow is
a member, and for which, given the restricted information they have of him,
such a situation would involve this kind of possibility, i.e., ‘giving lip back,’
which will give us the dilemma we need?

We need some kinds of rules and some kinds of categories - very simple
ones. Now one of the things we want to notice about that piece of
conversation I just quoted is that the interaction under which ‘giving lip’
might occur has, before ‘giving lip’ arises as a possibility, a formulation of
who it is that’s participating - a cop and a kid - where the ‘who it is’ is given
by the thing that’s happening. That is to say, this fellow was driving, and was
stopped by a policeman, as a possible violator. And one of the things that is
regular about that, is that the possibility of ‘giving lip back’ occurs in
situations which have been formulated by reference to some such categories.
For example, in interactions between high school teachers and students, in
interactions between parent and child. Where we have some first event that
takes place, done by other than the kid, which formulates the event as one for,
e.g., parent-child, teacher-student, policeman-teenage driver, etc. Those can
be things like a request or an order, which provide for the relevance of the
categories by virtue of the fact that the request or order ought to be accepted
by virtue of the categorial relationship between the parties. ‘Do the dishes,’
‘Put the cat out,’ are things for which the warrant is that a particular category
of person told you to do it, and thereby provides the relevance of that category
and whatever alternation there is to it. For example, ‘pulling you over’ is
something that ‘a cop’ does, and thereby provides the relevance of that he’s
‘a cop’ and you’re ‘a teenage driver.’

So we have some action that they do, that provides the relevance of a pair
of categories for the given interaction. Where, for one, that pair of categories,
under some formulation, provides that given the first action, the kid ought to
do some second action. Given the fact that you’ve been told “Put out the cat”
by someone who stands as ‘a parent,’ you ought to do it. Given the fact that
you’ve been told “Pull over” by someone who can be seen as ‘a cop,’ you
ought to do it. We have, then, a pair of categories, and a first activity.

There’s a general rule which seems to operate for any of the X categories
(cop, teacher, parent, etc.) and that alternation category, ‘kid’ (or whatever
variation on ‘kid’ that we would have; child, student, etc.). We could go
about formulating a series of alternatives. One could be, for example, there
might be a general rule that said ‘Don’t be insolent.’ Then there might be
another general rule, which need not be inconsistent with the first, which
would go ‘If somebody is nasty to you, then return it.’ The latter would deal
only with possible second statements. Then it would be the case that if some

446

Part III

person who ought to follow the first rule violates it, you can properly follow
the second rule. For anybody.

Now, whether or not there’s a general first rule like the one I’ve proposed,
is not terribly criterial for our purposes. What we need is to notice that there
is a rule which holds for any of those X categories and the alternative ones that
says: As between some pair of people, insults are non-reciprocal. That is to
say, whatever they do to you, you don’t do it in return. That rule, and the
prior second-statement rule I mentioned, can be in conflict. That is, that
second rule said, ‘If somebody does something like insulting, then you can
return it.’ And this one says, ‘For some people, if they do it, you oughtn’t.’
Of course they needn’t be in conflict in any given situation. Those people to
whom you should not return insults might not give them to you in the first
place.

As it happens, while the rule that says you can return an insult is quite
general, for certain groups, like kids - most particularly kids - it has achieved
a special status. It’s not simply permissable to do it, but, for one, as I earlier
mentioned, there’s a game in which that is central (the game called ‘tearing
down,’ also called ‘sounding,’ also called ‘playing the dozens’ and a variety of
other names). And secondly, the alternatives to a first event that consists of,
say, an insult, are both scorable; countable. That is to say, if you do return,
you show you’re not afraid. If you don’t, you show you’re chicken. So we
have, then, a situation where there’s a possible second event, and there are two
ways in which it could be dealt with. For kids, to either of those there’s a given
assessment: One is ‘ok,’ the other is ‘coward.’

The initiating event has, in the case of the data we’re considering, been
formulated as ‘having the car stopped for no good reason.’ It has occurred by
virtue of the fact that the driver of the car is a member of some kind of
alternation category, i.e., ‘teenage driver,’ ‘hotrodder,’ or some such thing as
that - a matter which, for purposes of stopping a car, is supposed to be
irrelevant, but which can be seen by both parties to have been the relevant
criteria for the interaction getting started. And which can be seen by the driver
of the car, specifically, to have been illegitimate.

And if it’s not a situation with cars, then the same formal situation will be
picked out by other kids to pose the same problem. Negro kids will pose it
by reference to their having done nothing; having been stopped by virtue of
the fact that they are Negro, and thereby people who are liable to do
something delinquent. Gang members in high schools formulate the matter
in just the same way. They focus on some situation in which the teacher tells
them to do something, where there’s no account available except that it was
done by virtue of the fact that they are obvious gang members - and of course
that could be obvious in a variety of ways.

Now they each take it that in such a situation, the first act is some sort of
illegitimate one, but an illegitimate one of a special sort: It’s one that, while
being illegitimate, nonetheless can have a guaranteed success for the one who
does it. In the case that we started with, there’s the fact that the cop can give
you non-moving violations. Fie oughtn’t to stop you unless he has good

Lecture 26

447

grounds. If he stops you, he can always find something that would have
added up to good grounds. That is to say, it won’t show up that he, in fact,
didn’t have any good grounds. He can give you a ticket, and that stands as
an account.

I’ve mentioned that methodology before, i.e., that an account can stand as
the account. It’s terribly powerful, and to see the character of its power one
need only to notice the following: Even though they attack it, the series of
questions they produce is really in its fashion, another case of the same thing.
That is to say, they ask him a series of questions which could add up to how
it is that such a one as he might have gotten to such a place as here. If they
find grounds, then they take it that those are the grounds.

Let me just give a citation while I have it in mind. There’s a paper by Carl
Werthman, ‘Delinquents in schools, a test for the legitimacy of authority,’
Berkeley Journal of Sociology (1964), pages 39-60. In there can be found a
whole bunch of parallel cases; enormous amounts of data in that paper.
There’s more data in that 20 pages, than in just about all the publications on
juvenile delinquency put together.

So: These various people in authority do something to that person who
they formulate as ‘a kid,’ something which is illegitimate. And it is specifically
illegitimate in the sense of picking up some aspect of the categorized state of
that person (the ‘kid’) which is supposedly irrelevant, and using that as the
occasion for doing that thing - stopping him, telling him to do something,
etc.

The matter is a little bit nicer than that, in that kids take it that cops,
teachers, etc., engage in an act like that with the specific aim of putting the
kid into a position that has no good solution. That is to say, they don’t just
happen to pick up a kid because they feel like picking up a kid, knowing it’s
illegitimate, but they pick up a kid knowing it’s illegitimate, and knowing the
kid sees it as illegitimate, and knowing that the sort of kid they pick up is put
into a position of having to make some determination of what it is he’s going
to do. That is, either be submissive or ‘give lip.’

And one possible reason that they do this is, given the fact that the
alternatives are assessable by other kids - if the kid ‘gives lip’ then he’s done
something that other kids say is fine, if not, then he’s done something that
makes him a coward - then, when these things are done in public, they
provide, e.g., the cops with a chance to show some tough guy to be not so
much. Where, for the kid, the alternative is if he does ‘give lip’ he’s going to
get arrested, although he’ll stand well with his buddies, and if he doesn’t, he
might not get arrested, but he’s been shown up in public.

In schools, the thing is used - from the kids’ point of view - as a way of
getting them out of school. Teachers know that the kids systematically will
respond by ‘giving lip back’ - the gang members, that is, who are in any
event a pain to the teacher. They’ll force the situation. And what the kid does
provides an occasion for having him kicked out of school. Now it’s beautifully
generative in its fashion, in that if you have a bunch of kids in the same gang,
then you need only get one of them who’s willing to stand up to the teacher,

448

Part III

for all of them to eventually get kicked out. That one stands up, he gets
kicked out, each of the others have their turns and have to show they’re
equivalent, or else. It’s not, then, for one, a chance matter that kids get kicked
out of school in Negro slums at the rate that they do.

Just to run through it again, then: What we have to have is a mechanism
of the following sort, for it to have a kind of guaranteed working. You need
somebody who’s not merely ‘just a kid,’ but who stands in some observable
status as a special kid - ‘hotrodder,’ ‘gang member,’ various other such
things. And you need somebody who stands as a member of one of several
special categories by reference to that kid - ‘cop,’ ‘teacher,’ ‘parent,’ etc. Then
you need an occasion when some activity can be done by somebody who can,
by that activity, formulate themselves as in one of these special categories
towards that kid. That activity will be some ‘first action,’ done by reference
to the special status of the kid, where that special status is specifically
irrelevant to doing the first action properly; that is to say, e.g., he shouldn’t
be stopped because of that status. That is the first move. And it ought to be
done in public, where ‘in public' means that at least some members are
present of the same category that this kid is in, so that he has relevant that
category’s set of rules and its consequences.

Those rules would be relevant in any event, under a situation of an
insult - it doesn’t have to be that a cop does such a thing. If any kid walks
by a bunch of such kids and makes a remark to them, then these things
should be operative. But if any kid does it, he can perfectly well be given back
the insult without any such systematic situation operating as does when an
adult does it - and particularly an adult in one or another authority status.

Now in part we have a question of why is it expectable that anybody in
such special categories would have that happen to them? Well, in the first
instance, as we discussed a while back, those categories are set up specifically
by the members with such an aim in mind. That is, they are set up as an
affront to authority, in order to set up a game, and to announce the fact that
we are hotrodders playing the game, against you who are supposedly
controlled by the fact that you can’t consider that when you decide whether
we’re speeding or not - w'here it’s part of our business to be specially
noticeable.

Under those sorts of circumstances, a question like ‘you give lip back to
everybody’ in this sequence, can provide as one of its further questions that
you’ve been kicked out of school. The alternative answers are tremendously
criterial as to what sort of kid he is, if the alternative answers are to be
believed. And note the kind of curious fact that only one of them would be
non-believed, and that is ‘yes.’ If he said ‘no,’ he’s presumably not lying. If
he says ‘yes,’ he may be lying. (That gives us a situation which may be similar
to saying to the question "How are you feeling?” “Fine.” Said to some class
of persons, it may be a lie.) If it is the case that only ‘yes’ would be heard as
a lie, then, that there is a norm for kids to ‘give lip back’ under such
circumstances gets a rather nice way of being seen. That is to say, one takes
it that one would only lie to another kid in the direction that involves

Lecture 26 449

conformity with the norms that hold between kids, and not the other way
around.

Now I just want to notice one thing, and it’s extremely tough to give a
decent formulation of. The term ‘bad’ can gain a lot of its power with
reference to dilemmas like this, by virtue of the fact that it can have two ways
of being heard. One is ‘bad,’ and one is “bad,” in quotes. Where “bad” in
quotes means ‘bad’ by reference to adult considerations, and ‘bad’ means by
reference to our considerations. Now that could be just in general, where a
quotative use of such terms might refer to that use of it which would be made
by some category that stands in alternation to ours. And the non-quotative
use could stand by reference to our use. So if you do one thing you’re ‘bad,’
and if you do the other thing you’re “bad.” What we want to find are some
activities which hold for anybody, not formulated by reference to, say, ‘kids’
or ‘hotrodders,’ in which that systematic situation for ‘bad’ holds for some
action that they can expectably all face.

Lecture 27

A mis-hearing (“a green?"); A taboo

on hearing

Roger: Are you just agreein’ because you feel you want to uh . . .

Jim : Hm?

Roger : You just agreeing?

Jim : What the hell’s that

Al : It’s // Agreeing?

Roger : Agreeing.

Jim : Agreen?

Roger : Yeah.

r Al : With us. Just going along with us.

'-Ken: Agreein’.

Jim : No.

What I want to consider now is a little of what may be involved in the
troubles that are present in this segment. Let’s assume for the moment that
Jim’s failure in the first instance is a mis-hearing, or the inability to hear what
he’s being asked. We can do that initially by virtue of the fact that, for one,
they take it that that’s what the trouble is, and also that he uses a way of
marking his trouble which has that use in the first place; that is, the question
“Hm?” which is a way of saying ‘What did you say?’

Then we can ask why should he mis-hear that or not hear what it is that’s
being asked? I suppose one could say, well, maybe it just wasn’t clear. To that,
there are a variety of returns. One of them is that it was heard by somebody
other than the one who said it i.e., “agreeing” is volunteered by another than
the one who asked the initial question. And when the item is repeated a series
of times, he still doesn’t get it. So even if it wasn’t clear in the first instance
that “agreeing” was said, he might have picked it up after a bit, but he
doesn’t. We want to ask why he doesn’t, and what sort of question it is, that
it involves such a response to it.

What that involves us in asking is, how is it that to some extent the fellow
who’s being questioned is attending the questions that he’s being asked? Is he
simply dealing with them as singular events - one question, whatever it is,
another question, whatever it is? Or is it the case, perhaps, that there’s some
relevance to the fact that he got a series of questions, and perhaps also that the
series is introduced by what could be seen as a rather general problem that’s
being investigated by them (i.e., “Tell us all about yourself so we can find
something bad about you”), and further, that the series of questions can be
seen - with that pre-first question and perhaps even in part independently of

450

Lecture 21

451

it - to have common features. So that the entry of another question to that list
would be understood by reference to what sort of features had been present
in items on that list. That is, he’s able to detect a common property to the set
of questions, which he can use to monitor a next — to decide that it’s one of
them, and which it is. And that would involve that in some sense the numbers
(“First of all,” “Secondly,” etc.) are not incidental, but that there is a series
with a number attached to each, and that has some significance for deciding
what it is that would be said with one of those numbers, or as part of the
series.

He may, then, be coming to use some way of seeing a next item by virtue
of the previous ones. And what that might give him is something like the
following: ‘Are you a [something]?’ Where what follows ‘Are you a’ is some
membership categorization category that it turns out he doesn’t know.
Something similar, for example, to what’s later asked, “Are you a hood?”
He’s been getting questions that run, ‘Are you like a such-and-such,’ ‘Are you
one who got kicked out of school,’ ‘Do you give lip back,’ etc. ‘Are you an
underachiever. ’

Then, his problem for ‘Are you a [something]?’ is finding what that thing
is. He may hear something that’s incomplete, in the sense that he hears ‘Are
you a green . . .’? where that might be ‘a green what?’ Or he may hear that
‘a green’ is something that they know of and use, which he doesn’t know of.
Some technical term of sorts, some ‘in’ term of sorts - where it would be that
sort of thing by virtue of the prior set of items. He would apparently take it
that the questions have a kind of linear extension in which it’s perfectly
reasonable to use what has been developed in them to see what the next
would look like, where that in fact controls what he hears, so as to hear,
indeed, a puzzle, when he could hear something perfectly clear.

That matter could be one that has rather large theoretical interest, in
dealing with the detection of speech and its understanding. Insofar as the issue
has been investigated in any interesting way, it turns on questions of the way
puns are built, and such things as hearing or not the correct word where a
possible homonym is used. One of the nice places, I suppose, to investigate
the problem, concerns obscenities, and there’s an extremely lovely article by E.
R. Leach, ‘Animal categories and verbal abuse,’ in Lenneberg (ed.), New
Directions in the Study of Language. Among the things that’s pointed out is
that one can, in routine talk, use a set of words which are more or less
unprintable epithets, and they’re not heard that way at all.

Now the tabooing of the hearing of obscene homonyms is an extraordi-
narily interesting thing. To see how interesting it is, you need only consider
occasions when one has opened up the situation of obscenity, and then tried
to avoid the hearing of almost anything that could be heard as obscene as not
obscene. That’s the way you get a chaining of humor, in that once obscenity
has been done it takes rather a while to return to a non-hearing of the obscene
possibility of the words used.

If you consider that sort of thing, then one could begin to see the possibility
with ‘a green.’ And that is, that it would just not occur to him to listen to the

452

Part III

object; perfectly plain on the tape for example, and perfectly plain after there’s
been three or four repetitions of it, as being a single word and not two words,
one of which is ‘a’ and the other of which is some noun that he doesn’t know;
a membership noun of some sort. And given the fact that the context - that
is, the set of questions — are usable, we have, then, some very nice evidence
that those set of questions are used to monitor a next.

Now that next question involved a shift in the kinds of questions that have
been asked, and not at all an incidental shift. For one, it’s related nicely to
what his own procedures apparently are for listening to those questions, in
that it’s a kind of complement to them. What’s involved is, he’s been using
their questions to formulate a way of seeing what a next question would be,
i.e., he’s seen some pattern in the set of questions, and hasn’t, apparently,
attended to the possible observing of a pattern in his set of answers.

I say that the shift is not incidental because being prepared for such kinds
of shifts, attending to them not only so as to hear what they are, but in the
first place attending to them in the production of a set of statements which can
be subjected to that, is one of the fundamental things that takes place, and has
to get learned, to be able to do this sort of therapy. I’ll say a little bit about
that.

One of the things that tends to be assumed by new patients to
therapy - and even, for that matter, relatively wise patients from certain sorts
of therapy - is that the situation is extremely similar to, for example, classical
medical institutions, and what is involved is the following: There is some
trouble in the outside world, you come in and talk about it, and you might
get some advice or an understanding of what it is that your trouble is, out
there.

In some of the American reformulations of classical psychotherapy
associated with Harry Stack Sullivan and ‘the Washington school,’ the
so-called ‘interactionists,’ psychotherapy involves, in part, the notion that
what takes place in therapy is not merely just as relevant, but is also a part of
the world which can be examined itself for finding what the troubles are.
Where, for one, it would then be argued that discussions about the outside
world, your troubles in youth, etc., are not a way of getting a good
understanding of why you come to have the troubles, what they are, etc., but
a way of dealing with your current problems in an easier fashion than to
confront them directly, here and now. If it’s older and outside, it might be
easier to deal with. So what you have, then, is a situation in which any given
interchange can form what is treated as the topic for the next: ‘Why did you
say that?,’ where that clarification that can go on will permit an understanding
of what it is that your troubles are.

It’s a kind of shift, then, where one gets a piece of talk directed to the ‘why’
of a prior piece, where the ‘why’ is altogether local: ‘Why did you say that
now?’ And sets of motives are employed which have nothing especially to do
with the particular matters under consideration, except that the same issues
might be found to be operative in behaviors there, as well as here. In that
event, of course, the way in which one would have to be on guard - if one

Lecture 21

453

wants to be on guard - is enormously heightened. For one, the fact of an
attention to the ordering of the way you proceed within the session, and the
use of that to formulate diagnoses, sets tremendous constraints on trying to
hide in that room. Where, especially, there’s not to be any constraint on the
order of things introduced as explanations with regard to the things they
apparently explain. That is to say, something that might seem to be perfectly
trivial, nothing, never getting noticed, like, for example, the fact that
somebody gets up and goes to the bathroom, is treated as an orientation to the
group; something for which a full complement of motives can be assigned,
and an explanation can be sought, and it can be watched for its future
occurrence, and the rest. And that’s what’s done quite routinely.

What you have, then, is that they make this place into what they talk
about as ‘a regular society,’ and what that involves is that any motive that
motivates people, any known reasons that people have for doing whatever
they do, however large they are, are now to be treated as usable with respect
to any activity that they happen to do here, however otherwise minor it might
be. So there are, then, a set of rules of relevance that are completely swept
away, which have to do with, for example, the fit between some action and
some reason for it. That you went to the bathroom because at that moment
you had to go to the bathroom, would be a routinely usable explanation
which is just about irrelevant here. He has to learn about those shifts, and
learn what are the sorts of things that can be offered, and he would be
expected to deal with, as accounts of whatever it is he’s been doing here. In
that regard, then, if he took the prior set of questions as a diagnosis, and his
answers as permitting a diagnosis, then he’s altogether misconceived what’s
going to take place.

But there are a series of ways that this question, “Are you just agreeing?’’
is much alike to prior questions. One of the beauties of that list of questions
is that while it can come off as being surprizingly knowing about this fellow
they haven’t met, there’s an obvious answer to how they know that, which is
tremendously relevant to him, and that is, that they’re telling him about
themselves. What they're doing in that set of questions, then, is presenting
information as to how it is that someone comes here, which of course
characterizes from their point of view, how they come here. Then those
questions are more or less true statements about themselves, as they see
it.

In that regard, then, we could ask about, and he might eventually think of,
the issue especially nicely tied with the question “Are you just agreeing?,’’
which is, why should it be supposed that he would have so produced his
answers? Why that reason for his answering as he did, but by virtue of the fact
that they know that that’s the way they did it. And then it comes to have a
series of possibly nice properties. For one, you could ask, Well, why would
they ask a question like that? since if in fact he was doing that, he isn’t going
to say it. It’s a question that forces a ‘no’ answer, whether the answer is ‘yes’
or ‘no.’ There’s a fairly decent answer to that: It could be a warning. Which
is: ‘If what you’ve been doing is just playing along with us, you’d better well

454

Part III

realize we could expect that that’s what you’re doing, and know that that’s
what you’re doing, so you’re not going to get away with it.’

That might be quite a frightening warning, and such matters are regularly
enormously frightening. That is, one is put into a position where whatever one
says is not going to be taken for what one says. What I mean by that is, the
character of questions and answers as ‘tied,’ is most neatly found by virtue of
the following: If you want to quote an answer to a question, you can perfectly
well take the question and the answer and combine them in a factual
statement which would go, ‘He said he was kicked out of school once’ or
‘He’s been kicked out of school once,’ where there’s a question and he says
Yes.’ That is, one doesn’t quote ‘Yes.’ One takes it that he said whatever it
is the question asked for.

Now, that’s being changed around here. For one, it can be read as, simply,
‘For some reason or other, to a question in which we gave him a possible
answer, he agreed with the answer - and who cares whether what he said is
true.’ He’s here, he’s sick, and whatever it is that it was is irrelevant. The only
question is, what does he do here? So you get a suspension of that kind of
quotative use of answers, except to point to the fact that ‘You said the
following, and why did you say it?’ That could be fearsome, because one
might never know what they’re seeing in some answer that looks perfectly
plain, and for which you could offer all kinds of warrants if it came down to
it - that is, e.g., you could bring in your expulsion certificate. The ways, then,
in which one is going to be viewed, are possibly frightening because what you
have as your best resource for deciding what they might find out, is what
awful things you know about yourself.

But then, it isn’t exactly as if a warning of that sort is offered by the
therapist. It’s rather softer than that. He’s being told about a thing that they
know about by virtue of the fact that they tried to do it also. And again, then,
it’s like the prior set of questions. That is to say, it can be seen as intendedly
informative of how it is that they behave or behaved.

And these remarks can be tied back to my earlier point about taboos on
hearing, in that the shift that took place is, apart from the fact that it
introduces a rule about therapy, an altogether illegitimate thing to do,
otherwise. And one of the ways you deal with an altogether illegitimate thing
to do, is to not hear at all that it was done. If you could hear your mother
saying some extraordinarily obscene remark, one way that you deal with it is
to hear that she’s saying something perfectly clean. In this case, he couldn’t
find what ‘clean,’ connected, statement was being made, and couldn’t hear
that sequencing violation. One doesn’t, for comparative strangers, inquire
into their motives for dealing with one.

I used the term ‘strangers’ at that point, not altogether incidentally. There’s
a rather exquisite place in Freud’s article on first therapy sessions. Now, Freud
is a tremendous stylist as a writer. And in that article he’s talking about the
question of when interpretations should be offered. And he wants to insist that
they should not be offered too early. The way he makes his point, stylistically,
is, elsewhere he always talks about ‘doctor’ and ‘patient.’ At this point he says

Lecture 21 455

something like, 'You shouldn’t tell something very deep to a stranger.’ And
in the next sentence he goes back to talking about ‘patient.’

In any event, I take it that the locus of this mis-hearing is quite
formulatable and probably very recurrent. And then, that the character of the
boys’ activities as doing diagnosis, is to be found in that sort of shift, and not
especially in the set of questions that have otherwise been offered - as though
one could compile an interview chart with a set of answers that would explain
how come he’s here.